- Email: [email protected]
An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments
BIOTRI-00022; No of Pages 8 Biotribology xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Biotribology journal homepage: http://www.elsevier.com/locate/biotri
An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments Junqi Ding Unilever Research & Development, 40 Merritt Blvd., Trumbull, CT 06611, United States
a r t i c l e
i n f o
Article history: Received 2 November 2015 Received in revised form 18 November 2015 Accepted 18 November 2015 Available online xxxx Keywords: Tactile Stick–slip Bio-tribology Soft materials Surfactant Friction measurement
a b s t r a c t The clean rinse feel of personal wash products is one of the major technical drivers of consumer preference and usually is measured by in-vivo consumer studies in the consumer goods industry. We report here a custom-made apparatus based on friction and vibration measurements that can be correlated to consumer perceptions of clean rinse feel. The apparatus consists of a rotary stage powered by a motor which is controlled by a programmable controller, a long swiveling arm, an artificial finger, and an underwater sample stage. The artificial finger can adjust the applied normal force on the substrate. The sliding speed of the artificial finger is adjusted and monitored through a computer. Data acquisition software is triggered by the combination of the software and hardware. Four sensors including two normally mounted load cells, an underwater hydrophone and an accelerometer attached to the surface of the artificial finger are used to detect the normal forces, vibration and underwater sound, while the artificial finger slides over the underwater substrate surface and washes the products off from substrate. A friction coefficient can be derived from the data of two load cells and rinse profiles constructed as a function of sliding time. Rinse profiles are shown to be different with application of different cleansers and are correlated to consumer perceptions of slimy or squeaky-clean for different products. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Typical consumer use of a personal cleanser includes several steps: applying the product to a surface (e.g., skin surface), diluting the product and then lathering, rinsing, drying, etc. In addition to the cleaning function, the in-use sensory perception of personal cleansers is a major technical driver of consumer preference. Particularly, the tactile sensations experienced during the washing process are important in governing a consumer's overall assessment. Changes in the tactile characteristics at the different wash stages cue particular effects, such as rough/smooth, sticky/slippery, slimy/clean feel; and therefore, determine overall product liking. Washing involves complex natural material surfaces, changing surfactant concentration during rinse, and benefit deposition, etc., which are very difficult to control and reproduce in a laboratory. In the consumer goods industry, consumer study is the accepted means to evaluate a product. But interpretation of the results of these studies can be problematic, particularly with regard to emotional state and personal preference as opposed to real product effects. Personal preference is not only related to the product properties, but also influenced strongly by the persons culture background. For example, Asian consumers, especially in Japan, often desire a certain squeaky clean sensation during rinsing, which cues the perception that the product has been thoroughly
E-mail address: [email protected].
rinsed off. In other regions, the same tactile sensation is usually connected to the harsh or soap-like perceptions. Even for the product effect, the skin surface may differ strongly from subject to subject because of age, skin condition, or other factors. Even for a single subject, results may differ from day to day due to use of other products, changing of skin conditions, etc. Most importantly, the high cost, time investment, and legal requirement limit the application of the consumer study (invivo), especially in the product development stage when new ingredients are introduced. Therefore, a laboratory test method is desired to study the physical phenomena during using of personal wash products during use. If we isolate other factors of the washing procedure (for example, fragrance delivery), in this paper, we focus on the characteristics of skin contact during the wash motion. The washing experience strongly relies on contact sensing, for example, applying a product to skin, lathering or rubbing the product on skin surface, followed by rinsing the product off the skin surface and finally feeling the skin after wash. These steps involve heavily with haptic perception. Haptic perception, one of the fundamental areas of cognitive engineering, has advanced with significant contributions from research in biology, engineering, and psychology [1,2,3,4]. Haptic perception of materials and surfaces relies largely on touch, or tactile sensing. In general, tactile sensing is essential for many applications: textile quality, identification of surface imperfections, robotic, medicine, etc., and attracts researchers from different fields [5,6,7,8,9,10,11, 12,13,14,15].
http://dx.doi.org/10.1016/j.biotri.2015.11.003 2352-5738/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
2
J. Ding / Biotribology xxx (2015) xxx–xxx
Force and friction between skin surfaces play important roles in touch perception, with extensive literature devoted to the topic [16, 17,18,19,20]. Due to the complexity of the skin surface, and availability of human subjects, many researchers focus on skin-polymer surfaces to understand the skin tribological properties. Those studies are more relevant to texture evaluation [21,7,20]. Tactile sensation is difficult to separate from other sensory aspects, therefore, it should be considered from a multisensory perspective (for review, see ref. [22]). In texture perception, there is no solid consensus on whether vision and touch are an integrated system [23,24]. However, there is general agreement that auditory inputs are important for judgment of texture [25,26,27,28]. Since mechanoreceptors under the skin surface are very sensitive to vibratory stimuli and contribute to tactile perception [29,30,31,32,33,34,35,36,37,38,39,40,41,42], vibration induced by sliding friction is clearly a key to understanding tactile perception [43,44,45]. In Unilever experiments, we found that the skin vibration and sound emission from rubbing fingers and palm across skin surface is highly correlated with perceived skin condition. Through the combination of an acoustic instrument and consumer studies, we have found [46,47] that the squeaky clean perception is connected to the stick–slip phenomena which can be detected via some measurable parameters, such as the surface vibration, sound emitted from the surfaces or friction on the surfaces during sliding of two surfaces against each other. Better defining the relationship between these measurable physical parameters and consumer “clean” feel during use of personal wash products inspired development of the apparatus reported here. An apparatus that can produce reproducible results within a laboratory that correlate well to the in-use personal perception is highly useful. Most commercial available instruments are based on tribological measurements which correlate poorly to consumer in-use perceptions. We believe that the poor correlation is at least partly due to lack of dynamic acoustic information in the lab data. The present apparatus is related to a previous invention reported in a separate publication [48]; the earlier instrument was built to measure friction simultaneously with dynamic quantities such as accelerations, forces, and sound pressures resulting from light sliding contact over a soft material in air, much like a finger lightly touching a soft material. It can be used to measure the dynamic quantities for leave-on products while the applied product dries. It has a unique capability to measure adhesion between surfaces, which directly relates to the stickiness of a surface. The method can measure and distinguish adhesion between surfaces and overcomes the difficulties associated with measuring friction force between soft surfaces. Use of a pair of force transducers allows the measurement of
both the normal contact force and the tangential dissipative force, without interfering with the sliding process. The apparatus described here is similar but can be used underwater. It captures the vibration of a sliding artificial finger, underwater sound and loads, simultaneously. It is thus well suited to evaluate the rinsing profile of personal wash products, and dynamic signals strongly correlate to consumer's in-use perceptions of personal wash products. This paper details the design of the instrument, its computer control, the required data processing to interpret experiment and demonstration of how a soap base and syndet (synthetic detergent) base cleanser can be reproducibly discriminated. 2. Instrument and materials 2.1. The apparatus 2.1.1. Hardware A top view of the custom-built apparatus is shown in Fig. 1. The apparatus consists several major components: motor, rotary stage, swiveling arm, “artificial finger”, sample stage and water bath. A computercontrolled motor (Emerson NTE-320 motor with Epsilon Eb-205 Digital Servo Drive, Control Techniques., Inc., Eden Prairie, MN) drives a rotary stage through a time-belt which damps noise from the motor rotation. A long swiveling arm fixed to the rotary stage provides near-linear motion; motion is most linear when the arm is much longer than the length of the sample stage. The sample stage, which is immersed into water bath (~50 l), is connected to the table by two load cells (Model AL311 Mid, Range 1000 g, Honeywell, Columbus, OH) which sense the vertical loads with a dynamic range from DC to 300 Hz with the amplifier (Model GM, Honeywell, Columbus, OH) appropriately configured. The water bath is placed on a scissor lifting table (MSC Direct, Melville, NY) for easy height adjustment so the sample stage can be conveniently submerged into or taken out of the water without interfering with any sensors. Near the sample stage, a hydrophone (model 8103, Brüel & Kjær, Norcross, GA, see Fig. 2) is used to detect sound underwater by the moving “artificial finger” sliding across the stationary sample substrate. The hydrophone signal is conditioned by a Nexus Charge Signal Conditioning Amplifier (Brüel & Kjær, Norcross, GA). An accelerometer (352A24, PCB Inc., Depew, NY) mounted on the “artificial finger” just above water detects the vibration of the “artificial finger”. To prevent damage from splashed water, we wrapped the accelerometer in a TEFLON tape. The electromagnetic coil attached to the “artificial finger” makes it possible to adjust the finger load during rubbing of the substrate, and to lift up during reverse movement by reversing the voltage
Fig. 1. The top view of the instrument diagram. The right side is the motor and rotary stages which are connected by a time-belt. Two limit switches are fixed in both sides to guide the rotary movement limits. Two round heavy counter-weight metal blocks are on the top of swiveling beam with easy adjustment to balance the beam. The long beam transforms the rotary movement into nearly linear movement for the “artificial finger” on the sample stage. The photosensor is aligned with the position guider which was calibrated to two load cells connected to the sample stage. The above parts are fixed on an optical table. A water bath is put on a cart next to the optical table.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
J. Ding / Biotribology xxx (2015) xxx–xxx
Fig. 2. The close-up schematic view of the sample stage and the sensors. The sample stage is near the edge of and connected to the optical table with two load cells. It is located to immerse in water bath with hydrophone while the load cells and accelerometer can be kept above the water level.
applied to the coil. The longer the artificial finger is, the more force is needed to lift it. The limit of the lift force resulted from a maximum current which the electromagnetic coil can have constrains the length of the artificial finger, therefore, constrains the water depth of the submerged sample stage. To determine the contact position of the artificial finger with the substrate, a photosensor was mounted on the swiveling beam and a calibrated white plastic plate was mounted on the table. While the artificial finger is at Position A in Fig. 3, the photosensor is toward one edge of the white plastic plate. The output from the photosensor changes from 0v to 9v at this position (from low to high). When the artificial finger moves away from the other vertical post on the same stage (Position B in Fig. 3), the photosensor leaves the white plastic plate and its output switches from 9v back to 0v (from high to low). By analyzing the time between the two edges of the signal (from low to high and from high to low, respectively) from the photosensor with the distance from those two signal edges, we can immediately locate the exact position of the artificial finger as long as the swiveling speed keeps constant in this period. An electromagnetic relay system (USB-ERB24, Measurement Computing Corp., Norton, MA) is controlled by computer to switch between applying the “finger” loads (during the sliding) or lifting the “finger” (during the reverse movement). 2.1.2. Automation and software design Fig. 4 illustrates the control diagram for the instrument. On the top of the diagram, a computer with Microsoft Windows XP operating system controls the instrument with custom software while simultaneously
3
Fig. 4. Data flow and control diagram for the different components in the apparatus. The arrows indicate the direction of data flows.
hosting commercial data acquisition software (B&K Pulse system, Brüel & Kjær, Norcross, GA). The Pulse software interacts with the LAN Interface Type 7533 and Input/Output Module Type 3032A (both from Brüel & Kjær, Norcross, GA) which digitizes different signals from the instrument. Six channels in the Input/Output Module were used for two load cells (two channels), an accelerometer (one channel), a photosensor (one channel), a hydrophone (one channel), the position photosensor (one channel) and motion direction (one channel). The Pulse system does not have a separated input channel for an external trigger to start or stop the data acquisition. Instead, it has to take one analog input channel to set a condition for starting or ending the data acquisition in the software. Although this uses up one valuable channel for high speed data acquisition input, it also provides an option for automation. Epsilon Eb-205 Digital Servo Drive controls the motor moving speed, direction, and has several analog output channels which can provide voltage output. It can be programmed to output voltage synced with the motor motion direction. The voltage output is used to trigger the data acquisition. Software written in Microsoft Visual C ++ with WTL (Windows Template Library) [49] integrates the different mechanical parts of the instrument (motor, moving direction, limit switches, force for the finger) and automates data acquisition. This software controls the motor through the Emerson Epsilon Eb-205, and the power for the electromagnetic coils of “artificial finger” through the digital relays (USB-ERB24 relay) and programmable power supply (9120A, B&K Precision, Yorba Linda, CA). With the combination of the commercial software and custom-written software, different components are integrated and automation is achieved.
Fig. 3. The photographs of the apparatus. The left photo shows the whole view of the apparatus with labels for key components. The right photo shows sample stage immersed in water bath. The Points 1 and 2 are the positions for two load cells. The Points A and B are two positions to connect the posts which extend the sample stage down to the water bath. The black foam underwater is the neoprene form with Vitro-Skin on it. Point 3 is the contact position for the “finger” surface and the substrate surface. The surface wave was generated by the stick–slip events during rinse-off personal wash product.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
4
J. Ding / Biotribology xxx (2015) xxx–xxx
2.1.3. Friction coefficient The setup with two normal load cells creates a system which can be used to measure the normal load, and also derive the sliding friction and friction coefficient. Fig. 5 shows the schematic diagram of the sample stage. The diagram is an analog to the stage photo in Fig. 3. The positions 1 and 2 are the points for the two load cells (N1 and N2) which can measure the normal force. The positions A and B are the points connected to the extended posts which connect to the sample stage. Point 3 is the position where two surfaces contact each other during sliding. Since the sample substrates and the “artificial finger” are both soft materials, the contact surface might be a narrow stripe instead of a single point. In that case, Point 3 should be the middle point of the contact area to simplify the force analysis without loss of the generality. In each position, the forces in two directions are labeled such as F1f, F2f and Ff for the tangential force, Fn, N1 and N2 for the normal force. We did not consider the weight of the sample stage and the buoyant force since N1 and N2 can include both forces and will not affect our calculation. In addition, since the two load cells are not in the same vertical plane with the contact area, there should be forces toward us from the diagram in Fig. 5. It does not affect our force analysis in other directions. Here we show how to calculate frictional force and coefficient from the two normal load cell readings. Based on force balance in normal and tangential directions: F n ¼ N1 þ N2
ð1Þ
F f ¼ F1f þ F2f
ð2Þ
where Fn is the normal force applied on the substrate by the artificial finger; N1 and N2 are normal forces measured by two load cells; Ff is the friction generated by the artificial finger on the substrate; F1f and F2f are the forces in the horizontal direction at the points of two load cells. The moments at point A and point B must be balanced too. At Point A: N1 L1 þ F 1 f h1 þ F 2 f h1 þ F n x ¼ F f h þ N2 ðL þ L1 Þ
ð3Þ
At Point B: N1 ðL þ L1 Þ þ F 1 f h1 þ F 2 f h1 ¼ F f h þ F n ðL−xÞ þ N 2 L1
ð4Þ
Therefore, the friction generated by the artificial finger sliding across the substrate can be summarized by: Ff ¼
1 ððN1 þ N2 Þðx þ L1 Þ−N2 ðL þ 2L1 ÞÞ h−h1
ð7Þ
and the friction coefficient can be obtained: μ¼
1 N2 ðL þ 2L1 Þ ðx þ L1 Þ− N1 þ N2 h−h1
ð8Þ
N1, N2, and x are measurable data, and L, L1, h and h1 are fixed for the apparatus. 2.2. Materials Two commercial body wash products with very different in-use feel, Lux Soft Kiss and Lux Recarregue from Unilever, were tested. Eight personal cleanser prototypes with composition differences (different surfactants, different polymers, with or without oil, etc.) were also examined. The hardness of tap water from our lab in Trumbull, Connecticut, is usually around 55 ppm CaCO3. To adjust water hardness, CaCl2·2H2O (Mallinckrodt Baker, Inc., Phillipsburg, NJ) and MgCl2·6H2O (EMD Chemicals Inc., Gibbstown, NJ) were added to water. Both chemicals were purchased from Sigma Aldrich, MO. Water hardness was adjusted to 100 ppm CaCO3 with the atomic ratio of Ca to Mg (3:1). To mimic the softness of deep skin and firmness of the skin surface, two layers of substrates were used for both the substrates as shown in Fig. 3. The bottom layer (black layer in the photo) is a weather-resistant neoprene form (McMasterCarr, Robbinsville, NJ) with a self-adhesive to attach it to the metal sample stage and “artificial finger”. The neoprene foam is 2 in. (5.08 cm) wide, 3/8 in. thick (0.95 cm), and has a firmness of 5–9 psi (25% deflection). On the top of the neoprene form, we affixed a thin Vitro-Skin (IMS Inc., Portland, ME). The Vitro-Skin is cut to an 1″× 5″ (2.54 cm × 12.7 cm) rectangle shape and pre-treated with warm water (~ 34 °C), then bound to the neoprene foam with rubber bands. The Vitro-Skin has two sides: one with a smooth surface and another with a coarse surface. The smooth surface is used for contacting personal wash products. 3. Application examples
Rearrange Eq. (3): N1 L1 þ F n x ¼ F f h− F 1 f h1 þ F 2 f h1 þ N 2 ðL þ L1 Þ
ð5Þ
Put Eqs. (1) and (2) into Eq. (5): N1 L1 þ ðN 1 þ N2 Þx ¼ F f ðh−h1 Þ þ N2 ðL þ L1 Þ:
ð6Þ
Fig. 5. The schematic diagram for force balance. Points 1, 2, 3 and A, B are corresponding points in Fig. 3.
3.1. Wash-off procedure To prepare the test, the water bath is filled with 50 L of water adjusted to the test temperature of 34 °C and target hardness (100 ppm for the data reported in this paper), and lifted up to immerse the sample stage. The high heat capacity of 50 L of water is sufficient to keep the experimental temperature stable for the duration of testing. The initial position of the “artificial finger” is as shown in Fig. 1. The “artificial finger” is programmed to take a certain load by applying a controlled current through the magnetic coil inside it and the swiveling arm is set to a specified moving speed to slide the “artificial finger” across the substrate under water. During this period, the data acquisition system is activated to digitize loads applied to the substrates, and measure the sound pressure under water, the vibration of the “artificial finger” and the photosensor signal for the contact position. When the swiveling arm reaches the end of the first sweep (the left end reference to Fig. 3), a limit switch is depressed, the current through the magnetic coil in the “artificial finger” reverses and the “artificial finger” lifts to prevent the “artificial finger” crashing against the sample stage. The swiveling arm moves back to initial position where another limit switch is depressed. This completes one washing cycle. After 10 washing cycles baseline to condition the Vitro-Skin, the water bath is lowered to bring the substrates out of water. After the Vitro-Skin is dried with paper towel to remove excess water, 0.25 g of personal wash product
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
J. Ding / Biotribology xxx (2015) xxx–xxx
is applied to the Vitro-Skin. 0.5 mL water is applied to dilute the personal wash product. The product is spread over the surface of Vitro-Skin at minimal speed to avoid lather. Then the sample stage is re-immersed in the water bath and data is collected over a 40 washing cycle period. 3.2. Data reconstruction Data from each washing cycle are stored in a file which, as discussed above, includes data from 6 channels. Fig. 6 shows signals from a single washing cycle where stick–slip was observed. Data from 5 channels are
Fig. 6. The signals from different sensors within one washing cycle. In the figure, signal a is the gate signal from the photosensor. The rising edge corresponds to Position A in Fig. 3, the finger contact position, and the falling edge to Position B. Signal b and c are the dynamic loads from two load cells, and signal d is the sum of two loads in Signal b and c. Signals e and f are the “finger” vibration and underwater sound, respectively. The two outside dash lines are identified by the rising and falling edges in Signal a. Two inner dash lines are the border lines to extract the signals from the finger sliding across the substrate. The numbers in b are 5 stages for one washing cycle.
5
plotted since the sixth channel digitizes the motor direction signal which is used to trigger the data acquisition software and does not reflect the rinse properties. From top to bottom in Fig. 6, signals from the photosensor, load cell 1, load cell 2, total normal load, accelerometer and underwater sound are plotted, respectively. Fig. 6b to f are highly correlated since they are signals from different sensors for the same process. Fig. 6b, for example, records the whole moving finger from initial position (very right side outside the view in Fig. 3) to the end (the very left side outside of view in Fig. 3). There are 5 stages in the signal. Stage 1: the “finger” freely approaches to the sample stage and only sample stage weight is recorded by the load cells. Stage 2: the “finger” touches the edge of sample stage when the pivot changes the “finger” angle to apply the force to the substrate. Stage 3: the “finger” passes across the rubber band which is used to fast the Vitro-Skin to the neoprene foam substrate. Stage 4: the “finger” surface slides across the substrate and moves away from the load cell. In this stage, the average sensed load decreases but the total normal load average remains nearly constant (Fig. 6d). During this period, stick–slip events are clearly observed and the load jumps up and down with the “finger” vibrations. Stage 5: the “finger” slides against the substrate edge creates oscillation. The useful information from sliding surfaces is in Stage 4. Fig. 6a plots the gate signal from the photosensor; this data reduces the work to extract the useful information. The rising edge corresponds to the “finger” at position A in Fig. 5, x = 0 in Eq. (8). The falling edge corresponds to the “finger” at position B in Fig. 5, x = L in Eq. (8). Based on the two edges in Fig. 6a, the unnecessary signals for all channels can be cut out automatically; however, the signal cannot be simply cut within the gate signal because the rubber bands used to hold the substrates create vibrations which can provide false signals. The rubber band is a little higher than the Vitro-Skin surface, therefore, the “finger” needs more time to only contact Vitro-Skin surface. It is advisable to cut an additional portion of the data sequence from both ends of the gated signal to get rid of edge and rubber band effects. In practice, about 1/6 of the whole gated range from both ends is cut off (between the two dash lines in both ends in Fig. 6). After cutting out signals from both ends, a short data sequence per channel remains. Fig. 7 lists three plots from total normal load (Fig. 7a, N1 + N2), “finger vibration” in term of acceleration (Fig. 7b) and underwater sound pressure (Fig. 7c). All three are highly correlated. The total normal load shows the clean signal of the force from the jumping finger generated by the stick–slip events. The vibration of the “finger” detected by the accelerometer tracks the “finger” jumping. Due to the applied force from the magnetic coils inside the pivot of the “finger”, the jumping of the “finger” is asymmetric. The amplitude of both normal load and “finger” vibration reflects the strength of the stick–slip events. Comparing to other signals, the underwater sound is noisier due to the ripple effect from the water. The stick–slip events from two surfaces can generate observable water waves as shown in the photograph in Fig. 3. It is actually sufficient to look at just the total normal load signal to find the stick–slip events. Each washing cycle creates useful signals which last more than 0.8 s at the swiveling speed of 33 mm/s. At lower speed, one obtains longer data sequences per washing cycle, and vice versa. In the real experiment, it might take about 20 s for the artificial finger to complete one cycle at the swiveling speed of 33 mm/s. For a simple application, we take 40 washing cycles to make sure the surfactants in the personal wash products are rinsed off. It helps visually to plot the extracted signals from each washing cycle together. Fig. 8 shows an example. In each plot, the x-axis is the reconstructed timeline which is accumulated time from the first to the last washing cycles by only considering the time taken for the useful signals. For example, the first washing cycle yields the first 0.8 s of useful time, the 2nd washing cycle's 0.8 s of useful time are put just behind the signal from first washing cycle, etc. The plots in Fig. 8 illustrate the rinse profiles from two personal wash products with distinct in-use feels. Lux Recarregue (Figs. 8a and 8b) is a syndet base body wash product which provides a soft and
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
6
J. Ding / Biotribology xxx (2015) xxx–xxx
whether it has stick–slip events, how strong the stick–slip is, how long the smooth region (slimy) is, whether the stick–slip is regular or irregular, the frequency of the stick–slip, and how the friction coefficient changes during rinse, etc. Analysis of all these parameters is beyond the scope of this paper, however, we illustrate how one of these parameters, the strength of stick–slip events, defined as the standard deviation of the total normal force in each washing cycle, can be used. Consider the standard deviation of the total normal force xi in each washing cycle: vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u n u1 X ðxi −xÞ2 : σ ¼t n i¼1
ð9Þ
This quantifies the amount of variation in the total normal force sensed by the two load cells. The bigger the standard deviation is, the stronger stick–slip is. Fig. 9 plots the standard deviation of the total normal force for 40 washing cycles from 8 different personal cleanser prototypes and two commercial products, Lux Soft Kiss and Lux Recarregue. Clearly, different cleansers have different behaviors in wash-off. Sample 7, 8 and Lux Soft Kiss have very high standard deviation after the initial low standard deviation which indicates the strong stick–slip after several washing cycles although Sample 7 shows slightly less stick–slip after the initial strong stick–slip. These samples provide clean rinse feel during wash after the initial “slimy” feel. Samples 2, 4, and 5, and Lux Soft Kiss, on the other hand, show a minimal standard deviation for all 40 washing cycles. Those samples create a very slimy feel during rinse. Interesting products are Samples 1, 3, and 6 which have more washing cycles of low standard deviation and generate intermediate stick–slip. The results provide rich information of measurable parameters connecting to formulations, benefit deposition and consumer perception. Generally, the new apparatus is a powerful tool to understand the relationship between consumer perception and measurable physical parameters. 3.4. Friction coefficient Fig. 7. The extracted signals from the finger sliding the sample substrate. The “squeaky” region is shown here from one washing cycle. Signal a is the total normal load from the sum of two load cells. Signal b is the “finger” vibration detected by the accelerometer. Signal c is the underwater sound pressure detected by hydrophone.
smooth rinse feel during wash. The major surfactants in Lux Recarregue are sodium lauryl ether sulfate (SLES) and cocamidopropyl betaine (CAPB). Lux Soft Kiss (Figs. 8c and 8d), by contrast, is a soap base (blend of fatty soaps with different chain lengths) body wash which provides a clean rinse feel during wash. For Lux Recarregue, both total normal load and “finger” vibration do not show stick–slip for all 40 washing cycles. In contrast, for Lux Soft Kiss, both total normal load and “finger” vibration show two distinct regions in 40 washing cycles. The first two washing cycles do not have any stick–slip. The sliding movement of the “finger” across the substrate is very smooth; this corresponds to the slimy region. At this stage, there are enough surfactant on the skin surface (for a consumer) or on the Vitro-Skin surface (as in the apparatus reported here) to generate very slippery surfaces. Starting from the 3rd washing cycle, a strong stick–slip commences; this corresponds to the consumer's reported “squeaky-clean” feel when using his/her fingers to slide across skin surface. Generally, the apparatus results are highly correlated to in-use perceptions. 3.3. Simplification Fig. 8 provides visual plots for smooth or stick–slip movements of sliding “finger” which can be used to predict in-use perceptions for personal wash products. These plots have extremely large amount of data. Simplification of the raw data should be considered to extract meaningful parameters. There are many ways to analyze the raw data. Taking the total normal force, for example, we can analyze the data to determine:
The apparatus uses two load cells which can be used to derive friction coefficient indirectly. Fig. 10 plots the derived friction coefficients of 6 cleanser prototypes and Lux Recarregue. Lux Soft Kiss and Samples 7, 8 shown in Fig. 9 do not have meaningful friction coefficient due to the strong stick–slip during rinse. All cleansers have very low friction coefficients for the first two washing cycles because the initial local high concentration of surfactants act as lubricants. All samples except Samples 3 and 6 retain very low friction coefficients for all 40 washing cycles. The slight difference among those samples is likely not significant. Samples 3 and 6 however have significantly different trends from other samples. Sample 6 has basically the same surfactant system as Lux Recarregue with an additional hydrophobic polymer. A hydrophobic polymer with high molecular weight (typically C2–C10 polyalkenes such as polybutene here) in Sample 6 tunes the rinse and friction coefficient profile. After the surfactants on the substrate surfaces is washed off, the effect of the polymer dominates the friction coefficient during the rest of washing which results a sudden rise in friction coefficient. Sample 3 contains a surfactant combination with the major surfactant of Sodium Cocoyl Isethionate (SCI), a gentle surfactant based on coconut fatty acids. It has an intermediate friction properties. Friction coefficient changes during rinse can be used to understand the effect of different raw materials, product structuring and formulation, etc., therefore, and guide product formulation to meet consumers' expectations without incurring the much more significant costs of a consumer study. 4. Summary We reported the design and control of a custom-built apparatus to measure friction and vibration profiles during simulated washing with personal wash products. We described how to automate data collection,
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
J. Ding / Biotribology xxx (2015) xxx–xxx
7
Fig. 8. The rebuilt rinse profiles for two LUX cleansers which have different in-use feelings. The left two plots (a and c) are the total normal load, and the right two plots (b and d) are the “finger” vibration, from 40 washing cycles. The top two plots are the rinse profiles from LUX Recarregue and the bottom two plots are the rinse profiles from LUX Soft Kiss. In c and d, the stick–slip events generate the big oscillations in the signal from the 3rd washing cycle. To observe the detail of the stick–slip in each washing cycle, refer Fig. 7.
and cut unnecessary data to limit bandwidth while retaining all necessary information to reconstruct rinse profiles. We have shown that these rinse profiles differ with different cleansers in ways that have straightforward connections to the consumer tactile experience. For the soap base cleansers such as the Lux Soft Kiss product, the instrument indicates a low amplitude in the finger vibration, low noise of underwater sound, and constant normal force, during the first two washing cycles, corresponding to a very short slimy region. Following the slimy region, the soap system generates stick–slip events which trigger the strong finger vibration, higher underwater noise, and large normal
Fig. 9. The standard deviation of total normal force from different personal cleansers.
force variation. For the syndet base cleanser such as Lux Recarregue product, stick–slip events generally do not occur through the whole rinse period. The results correlate with the clean-rinse feel reported by consumers. Tuning the cleanser formulation with different techniques affects the friction coefficient during rinsing. The apparatus appears to be a promising tool to study the in-vitro rinse properties of cleansers.
Fig. 10. The friction coefficient of different personal cleansers indirectly measured from two load cells with the “artificial finger” apparatus in term of washing progress. For two samples with strong stick–slip (samples 7 and 8) shown in Fig 9, the friction efficient is not meaningful due to the discontinuous contact of two surfaces and is not shown here.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
8
J. Ding / Biotribology xxx (2015) xxx–xxx
Acknowledgments The author acknowledges the support from and helpful discussions with Larry Wilen, KP Ananth and Albert Post. The early collaborations with Akay Adnan, Blair Echols, Anne Dussaud, Frank Meyer and Alex Lips which led to the current work are also highly appreciated. Daniel Rodriguez helped to create several drawings. The author would like to thank Heidi Warriner for proofreading of the final draft. References [1] M.H. Lee, Tactile sensing: new directions, new challenges, Int. J. Robot. Res. 19 (7) (2000) 636–643. [2] R. Dahiya, P. Mittendorfer, M. Valle, G. Cheng, V. Lumelsky, Directions toward effective utilization of tactile skin: a review, IEEE Sensors J. 13 (11) (2013) 4121–4138, http://dx.doi.org/10.1109/JSEN.2013.2279056. [3] S. Stassi, V. Cauda, G. Canavese, C. Pirri, Flexible tactile sensing based on piezoresistive composites: a review, Sensors (Switzerland) 14 (3) (2014) 5296–5332, http://dx.doi.org/10.3390/s140305296. [4] P. Shull, D. Damian, Haptic wearables as sensory replacement, sensory augmentation and trainer - a review, J. Neuroeng. Rehabil. 12 (1) (2015)http://dx.doi.org/ 10.1186/s12984-015-0055-z. [5] R. Lundström, L. Burström, Mechanical impedance of the human hand–arm system, Int. J. Ind. Ergon. 3 (3) (1989) 235–242, http://dx.doi.org/10.1016/01698141(89)90023-1. [6] H. Limerick, D. Coyle, J. Moore, The experience of agency in human-computer interactions: a review, Front. Hum. Neurosci. (2014) 8, http://dx.doi.org/10.3389/fnhum. 2014.00643 (AUG). [7] X. Liao, Y. Li, J. Hu, X. Wu, Q. Li, A simultaneous measurement method to characterize touch properties of textile materials, Fibers Polym. 15 (7) (2014) 1548–1559, http://dx.doi.org/10.1007/s12221-014-1548-2. [8] A. Klöcker, M. Wiertlewski, V. Théate, V. Hayward, J.L. Thonnard, Physical factors influencing pleasant touch during tactile exploration, PLoS ONE 8 (11) (2013)http://dx.doi.org/10.1371/journal.pone.0079085. [9] R. Samra, M. Zadeh, D. Wang, Design of a tactile instrument to measure human roughness perception in a virtual environment, IEEE Trans. Instrum. Meas. 60 (11) (2011) 3582–3591, http://dx.doi.org/10.1109/TIM.2011.2161149. [10] C. Lucarotti, C. Oddo, N. Vitiello, M. Carrozza, Synthetic and bio-artificial tactile sensing: a review, Sensors (Switzerland) 13 (2) (2013) 1435–1466, http://dx.doi.org/10. 3390/s130201435. [11] A. Hussain, A. Malik, M. Halim, A. Ali, The use of robotics in surgery: a review, Int. J. Clin. Pract. 68 (11) (2014) 1376–1382, http://dx.doi.org/10.1111/ijcp.12492. [12] H. Zahouani, S. Mezghani, R. Vargiolu, T. Hoc, M. El Mansori, Effect of roughness on vibration of human finger during a friction test, Wear 301 (1–2) (2013) 343–352, http://dx.doi.org/10.1016/j.wear.2012.11.028. [13] A. Sawers, L. Ting, Perspectives on human-human sensorimotor interactions for the design of rehabilitation robots, J. Neuroeng. Rehabil. 11 (1) (2014)http://dx.doi.org/ 10.1186/1743-0003-11-142. [14] A. Zimmerman, L. Bai, D. Ginty, The gentle touch receptors of mammalian skin, Science 346 (6212) (2014) 950–954, http://dx.doi.org/10.1126/science.1254229. [15] T. Heed, V. Buchholz, A. Engel, B. Röder, Tactile remapping: from coordinate transformation to integration in sensorimotor processing, Trends Cogn. Sci. 19 (5) (2015) 251–258, http://dx.doi.org/10.1016/j.tics.2015.03.001. [16] B.N.J. Persson, Sliding Friction physical Principles and Applications, NanoScience and Technology, Springer-Verlag, Berlin Heidelberg, 2000http://dx.doi.org/10.1007/9783-662-04283-0. [17] M. Adams, B. Briscoe, S. Johnson, Friction and lubrication of human skin, Tribol. Lett. 26 (3) (2007) 239–253, http://dx.doi.org/10.1007/s11249-007-9206-0. [18] M. Adams, S. Johnson, V. Lefèvre, P. Lévesque, V. Hayward, T. André, J.L. Thonnard, Finger pad friction and its role in grip and touch, J. R. Soc. Interface 10 (80) (2013)http://dx.doi.org/10.1098/rsif.2012.0467. [19] S. Pasumarty, S. Johnson, S. Watson, M. Adams, Friction of the human finger pad: influence of moisture, occlusion and velocity, Tribol. Lett. 44 (2) (2011) 117–137, http://dx.doi.org/10.1007/s11249-011-9828-0. [20] S. Schreiner, M. Rechberger, J. Bertling, Haptic perception of friction — correlating friction measurements of skin against polymer surfaces with subjective evaluations of the surfaces' grip, Tribol. Int. 63 (2013) 21–28, http://dx.doi.org/10.1016/j. triboint.2012.08.017. [21] A. Takahashi, M. Suzuki, Y. Imai, Y. Nonomura, Tactile texture and friction of soft sponge surfaces, Colloids Surf. B: Biointerfaces 130 (2015) 10–15, http://dx.doi. org/10.1016/j.colsurfb.2015.03.055. [22] A. Gallace, C. Spence, The cognitive and neural correlates of “tactile consciousness”: a multisensory perspective, Conscious. Cogn. 17 (1) (2008) 370–407, http://dx.doi. org/10.1016/j.concog.2007.01.005.
[23] T. Whitaker, C. Simões Franklin, F. Newell, Vision and touch: independent or integrated systems for the perception of texture? Brain Res. 1242 (2008) 59–72, http://dx.doi.org/10.1016/j.brainres.2008.05.037. [24] Y.-C. Pei, T.-Y. Chang, T.-C. Lee, S. Saha, H.-Y. Lai, M. Gomez-Ramirez, et al., Crossmodal sensory integration of visual–tactile motion information: Instrument design and human psychophysics, Sensors (Switzerland) 13 (6) (2013) 7212–7223, http://dx.doi.org/10.3390/s130607212. [25] S. Soto-Faraco, G. Deco, Multisensory contributions to the perception of vibrotactile events, Behav. Brain Res. 196 (2) (2009) 145–154, http://dx.doi.org/10.1016/j.bbr. 2008.09.018. [26] V. Occelli, C. Spence, M. Zampini, Audiotactile interactions in front and rear space, Neurosci. Biobehav. Rev. 35 (3) (2011) 589–598, http://dx.doi.org/10.1016/j. neubiorev.2010.07.004. [27] G. Deco, R. Romo, The role of fluctuations in perception, Trends Neurosci. 31 (11) (2008) 591–598, http://dx.doi.org/10.1016/j.tins.2008.08.007. [28] J. Foxe, Multisensory integration: frequency tuning of audio-tactile integration, Curr. Biol. 19 (9) (2009) R373–R375, http://dx.doi.org/10.1016/j.cub.2009.03.029. [29] J. Löfvenberg, R. Johansson, Regional differences and interindividual variability in sensitivity to vibration in the glabrous skin of the human hand, Brain Res. 301 (1) (1984) 65–72, http://dx.doi.org/10.1016/0006-8993(84)90403-7. [30] H. Hämäläinen, T. Järvilehto, K. Soininen, Vibrotactile atonal interval correlated with activity in peripheral mechanoreceptive units innervating the human hand, Brain Res. 333 (2) (1985) 311–324, http://dx.doi.org/10.1016/0006-8993(85)91585-9. [31] G. Gescheider, S. Bolanowski Jr., R. Verrillo, Vibrotactile masking: effects of stimulus onset asynchrony and stimulus frequency, J. Acoust. Soc. Am. 85 (5) (1989) 2059–2064. [32] G. Gescheider, S. Bolanowski Jr., R. Verrillo, D. Arpajian, T. Ryan, Vibrotactile intensity discrimination measured by three methods, J. Acoust. Soc. Am. 87 (1) (1990) 330–338. [33] J. Bell, S. Bolanowski, M. Holmes, The structure and function of pacinian corpuscles: a review, Prog. Neurobiol. 42 (1) (1994) 79–128, http://dx.doi.org/10.1016/03010082(94)90022-1. [34] M. Hollins, E. Roy, Perceived intensity of vibrotactile stimuli: The role of mechanoreceptive channels, Somatosens. Mot. Res. 13 (3–4) (1996) 273–286. [35] A. Brisben, S. Hsiao, K. Johnson, Detection of vibration transmitted through an object grasped in the hand, J. Neurophysiol. 81 (4) (1999) 1548–1558. [36] K. Johnson, T. Yoshioka, F. Vega Bermudez, Tactile functions of mechanoreceptive afferents innervating the hand, J. Clin. Neurophysiol. 17 (6) (2000) 539–558. [37] G. Gescheider, S. Bolanowski, K. Hardick, The frequency selectivity of informationprocessing channels in the tactile sensory system, Somatosens. Mot. Res. 18 (3) (2001) 191–201, http://dx.doi.org/10.1080/01421590120072187. [38] K. Johnson, The roles and functions of cutaneous mechanoreceptors, Curr. Opin. Neurobiol. 11 (4) (2001) 455–461, http://dx.doi.org/10.1016/S09594388(00)00234-8. [39] G. Gescheider, S. Bolanowski, J. Pope, R. Verrillo, A four-channel analysis of the tactile sensitivity of the fingertip: frequency selectivity, spatial summation, and temporal summation, Somatosens. Mot. Res. 19 (2) (2002) 114–124, http://dx.doi.org/10. 1080/08990220220131505. [40] G. Gescheider, S. Bolanowski, R. Verrillo, Some characteristics of tactile channels, Behav. Brain Res. 148 (1–2) (2004) 35–40, http://dx.doi.org/10.1016/S01664328(03)00177-3. [41] B. Güçlü, S. Bolanowski, Vibrotactile thresholds of the non-pacinian i channel: I. methodological issues, Somatosens. Mot. Res. 22 (1–2) (2005) 49–56, http://dx. doi.org/10.1080/08990220500084719. [42] B. Güçlü, S. Bolanowski, Vibrotactile thresholds of the non-pacinian i channel: II. predicting the effects of contactor location on the phalanx, Somatosens. Mot. Res. 22 (1–2) (2005) 57–68, http://dx.doi.org/10.1080/08990220512331387971. [43] A. Akay, Acoustics of friction, J. Acoust. Soc. Am. 111 (4) (2002) 1525–1548, http:// dx.doi.org/10.1121/1.1456514. [44] R. Fagiani, F. Massi, E. Chatelet, Y. Berthier, A. Akay, Tactile perception by friction induced vibrations, Tribol. Int. 44 (10) (2011) 1100–1110, http://dx.doi.org/10.1016/j. triboint.2011.03.019. [45] R. Fagiani, F. Massi, E. Chatelet, J. Costes, Y. Berthier, Contact of a finger on rigid surfaces and textiles: friction coefficient and induced vibrations, Tribol. Lett. 48 (2) (2012) 145–158, http://dx.doi.org/10.1007/s11249-012-0010-0. [46] J. Ding, A. Dussaud, A. Lips, A novel method for characterizing tactile perception, Abstracts of Papers, 227th ACS National Meeting, Anaheim, CA, United States, March 28-April 1 2004, p. 2004 COLL-420. [47] A. Dussaud, J. Ding, A. Lips, Method and system for characterizing tactile perception, U.S. Pat. Appl. Publ. (2005) 20050106103 A1 (US 20050106103 A1). [48] A. Akay, B. Echols, J. Ding, A. Dussaud, A. Lips, Measurement of sound, vibration and friction between soft materials under light loads, Wear 276–277 (2012) 61–69, http://dx.doi.org/10.1016/j.wear.2011.12.004. [49] M. Dunn, WTL for MFC programmers, 2005http://www.codeproject.com/Articles/ 3867/WTL-for-MFC-Programmers-Part-II-WTL-GUI-Base-Class.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
Contents lists available at ScienceDirect
Biotribology journal homepage: http://www.elsevier.com/locate/biotri
An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments Junqi Ding Unilever Research & Development, 40 Merritt Blvd., Trumbull, CT 06611, United States
a r t i c l e
i n f o
Article history: Received 2 November 2015 Received in revised form 18 November 2015 Accepted 18 November 2015 Available online xxxx Keywords: Tactile Stick–slip Bio-tribology Soft materials Surfactant Friction measurement
a b s t r a c t The clean rinse feel of personal wash products is one of the major technical drivers of consumer preference and usually is measured by in-vivo consumer studies in the consumer goods industry. We report here a custom-made apparatus based on friction and vibration measurements that can be correlated to consumer perceptions of clean rinse feel. The apparatus consists of a rotary stage powered by a motor which is controlled by a programmable controller, a long swiveling arm, an artificial finger, and an underwater sample stage. The artificial finger can adjust the applied normal force on the substrate. The sliding speed of the artificial finger is adjusted and monitored through a computer. Data acquisition software is triggered by the combination of the software and hardware. Four sensors including two normally mounted load cells, an underwater hydrophone and an accelerometer attached to the surface of the artificial finger are used to detect the normal forces, vibration and underwater sound, while the artificial finger slides over the underwater substrate surface and washes the products off from substrate. A friction coefficient can be derived from the data of two load cells and rinse profiles constructed as a function of sliding time. Rinse profiles are shown to be different with application of different cleansers and are correlated to consumer perceptions of slimy or squeaky-clean for different products. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Typical consumer use of a personal cleanser includes several steps: applying the product to a surface (e.g., skin surface), diluting the product and then lathering, rinsing, drying, etc. In addition to the cleaning function, the in-use sensory perception of personal cleansers is a major technical driver of consumer preference. Particularly, the tactile sensations experienced during the washing process are important in governing a consumer's overall assessment. Changes in the tactile characteristics at the different wash stages cue particular effects, such as rough/smooth, sticky/slippery, slimy/clean feel; and therefore, determine overall product liking. Washing involves complex natural material surfaces, changing surfactant concentration during rinse, and benefit deposition, etc., which are very difficult to control and reproduce in a laboratory. In the consumer goods industry, consumer study is the accepted means to evaluate a product. But interpretation of the results of these studies can be problematic, particularly with regard to emotional state and personal preference as opposed to real product effects. Personal preference is not only related to the product properties, but also influenced strongly by the persons culture background. For example, Asian consumers, especially in Japan, often desire a certain squeaky clean sensation during rinsing, which cues the perception that the product has been thoroughly
E-mail address: [email protected].
rinsed off. In other regions, the same tactile sensation is usually connected to the harsh or soap-like perceptions. Even for the product effect, the skin surface may differ strongly from subject to subject because of age, skin condition, or other factors. Even for a single subject, results may differ from day to day due to use of other products, changing of skin conditions, etc. Most importantly, the high cost, time investment, and legal requirement limit the application of the consumer study (invivo), especially in the product development stage when new ingredients are introduced. Therefore, a laboratory test method is desired to study the physical phenomena during using of personal wash products during use. If we isolate other factors of the washing procedure (for example, fragrance delivery), in this paper, we focus on the characteristics of skin contact during the wash motion. The washing experience strongly relies on contact sensing, for example, applying a product to skin, lathering or rubbing the product on skin surface, followed by rinsing the product off the skin surface and finally feeling the skin after wash. These steps involve heavily with haptic perception. Haptic perception, one of the fundamental areas of cognitive engineering, has advanced with significant contributions from research in biology, engineering, and psychology [1,2,3,4]. Haptic perception of materials and surfaces relies largely on touch, or tactile sensing. In general, tactile sensing is essential for many applications: textile quality, identification of surface imperfections, robotic, medicine, etc., and attracts researchers from different fields [5,6,7,8,9,10,11, 12,13,14,15].
http://dx.doi.org/10.1016/j.biotri.2015.11.003 2352-5738/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
2
J. Ding / Biotribology xxx (2015) xxx–xxx
Force and friction between skin surfaces play important roles in touch perception, with extensive literature devoted to the topic [16, 17,18,19,20]. Due to the complexity of the skin surface, and availability of human subjects, many researchers focus on skin-polymer surfaces to understand the skin tribological properties. Those studies are more relevant to texture evaluation [21,7,20]. Tactile sensation is difficult to separate from other sensory aspects, therefore, it should be considered from a multisensory perspective (for review, see ref. [22]). In texture perception, there is no solid consensus on whether vision and touch are an integrated system [23,24]. However, there is general agreement that auditory inputs are important for judgment of texture [25,26,27,28]. Since mechanoreceptors under the skin surface are very sensitive to vibratory stimuli and contribute to tactile perception [29,30,31,32,33,34,35,36,37,38,39,40,41,42], vibration induced by sliding friction is clearly a key to understanding tactile perception [43,44,45]. In Unilever experiments, we found that the skin vibration and sound emission from rubbing fingers and palm across skin surface is highly correlated with perceived skin condition. Through the combination of an acoustic instrument and consumer studies, we have found [46,47] that the squeaky clean perception is connected to the stick–slip phenomena which can be detected via some measurable parameters, such as the surface vibration, sound emitted from the surfaces or friction on the surfaces during sliding of two surfaces against each other. Better defining the relationship between these measurable physical parameters and consumer “clean” feel during use of personal wash products inspired development of the apparatus reported here. An apparatus that can produce reproducible results within a laboratory that correlate well to the in-use personal perception is highly useful. Most commercial available instruments are based on tribological measurements which correlate poorly to consumer in-use perceptions. We believe that the poor correlation is at least partly due to lack of dynamic acoustic information in the lab data. The present apparatus is related to a previous invention reported in a separate publication [48]; the earlier instrument was built to measure friction simultaneously with dynamic quantities such as accelerations, forces, and sound pressures resulting from light sliding contact over a soft material in air, much like a finger lightly touching a soft material. It can be used to measure the dynamic quantities for leave-on products while the applied product dries. It has a unique capability to measure adhesion between surfaces, which directly relates to the stickiness of a surface. The method can measure and distinguish adhesion between surfaces and overcomes the difficulties associated with measuring friction force between soft surfaces. Use of a pair of force transducers allows the measurement of
both the normal contact force and the tangential dissipative force, without interfering with the sliding process. The apparatus described here is similar but can be used underwater. It captures the vibration of a sliding artificial finger, underwater sound and loads, simultaneously. It is thus well suited to evaluate the rinsing profile of personal wash products, and dynamic signals strongly correlate to consumer's in-use perceptions of personal wash products. This paper details the design of the instrument, its computer control, the required data processing to interpret experiment and demonstration of how a soap base and syndet (synthetic detergent) base cleanser can be reproducibly discriminated. 2. Instrument and materials 2.1. The apparatus 2.1.1. Hardware A top view of the custom-built apparatus is shown in Fig. 1. The apparatus consists several major components: motor, rotary stage, swiveling arm, “artificial finger”, sample stage and water bath. A computercontrolled motor (Emerson NTE-320 motor with Epsilon Eb-205 Digital Servo Drive, Control Techniques., Inc., Eden Prairie, MN) drives a rotary stage through a time-belt which damps noise from the motor rotation. A long swiveling arm fixed to the rotary stage provides near-linear motion; motion is most linear when the arm is much longer than the length of the sample stage. The sample stage, which is immersed into water bath (~50 l), is connected to the table by two load cells (Model AL311 Mid, Range 1000 g, Honeywell, Columbus, OH) which sense the vertical loads with a dynamic range from DC to 300 Hz with the amplifier (Model GM, Honeywell, Columbus, OH) appropriately configured. The water bath is placed on a scissor lifting table (MSC Direct, Melville, NY) for easy height adjustment so the sample stage can be conveniently submerged into or taken out of the water without interfering with any sensors. Near the sample stage, a hydrophone (model 8103, Brüel & Kjær, Norcross, GA, see Fig. 2) is used to detect sound underwater by the moving “artificial finger” sliding across the stationary sample substrate. The hydrophone signal is conditioned by a Nexus Charge Signal Conditioning Amplifier (Brüel & Kjær, Norcross, GA). An accelerometer (352A24, PCB Inc., Depew, NY) mounted on the “artificial finger” just above water detects the vibration of the “artificial finger”. To prevent damage from splashed water, we wrapped the accelerometer in a TEFLON tape. The electromagnetic coil attached to the “artificial finger” makes it possible to adjust the finger load during rubbing of the substrate, and to lift up during reverse movement by reversing the voltage
Fig. 1. The top view of the instrument diagram. The right side is the motor and rotary stages which are connected by a time-belt. Two limit switches are fixed in both sides to guide the rotary movement limits. Two round heavy counter-weight metal blocks are on the top of swiveling beam with easy adjustment to balance the beam. The long beam transforms the rotary movement into nearly linear movement for the “artificial finger” on the sample stage. The photosensor is aligned with the position guider which was calibrated to two load cells connected to the sample stage. The above parts are fixed on an optical table. A water bath is put on a cart next to the optical table.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
J. Ding / Biotribology xxx (2015) xxx–xxx
Fig. 2. The close-up schematic view of the sample stage and the sensors. The sample stage is near the edge of and connected to the optical table with two load cells. It is located to immerse in water bath with hydrophone while the load cells and accelerometer can be kept above the water level.
applied to the coil. The longer the artificial finger is, the more force is needed to lift it. The limit of the lift force resulted from a maximum current which the electromagnetic coil can have constrains the length of the artificial finger, therefore, constrains the water depth of the submerged sample stage. To determine the contact position of the artificial finger with the substrate, a photosensor was mounted on the swiveling beam and a calibrated white plastic plate was mounted on the table. While the artificial finger is at Position A in Fig. 3, the photosensor is toward one edge of the white plastic plate. The output from the photosensor changes from 0v to 9v at this position (from low to high). When the artificial finger moves away from the other vertical post on the same stage (Position B in Fig. 3), the photosensor leaves the white plastic plate and its output switches from 9v back to 0v (from high to low). By analyzing the time between the two edges of the signal (from low to high and from high to low, respectively) from the photosensor with the distance from those two signal edges, we can immediately locate the exact position of the artificial finger as long as the swiveling speed keeps constant in this period. An electromagnetic relay system (USB-ERB24, Measurement Computing Corp., Norton, MA) is controlled by computer to switch between applying the “finger” loads (during the sliding) or lifting the “finger” (during the reverse movement). 2.1.2. Automation and software design Fig. 4 illustrates the control diagram for the instrument. On the top of the diagram, a computer with Microsoft Windows XP operating system controls the instrument with custom software while simultaneously
3
Fig. 4. Data flow and control diagram for the different components in the apparatus. The arrows indicate the direction of data flows.
hosting commercial data acquisition software (B&K Pulse system, Brüel & Kjær, Norcross, GA). The Pulse software interacts with the LAN Interface Type 7533 and Input/Output Module Type 3032A (both from Brüel & Kjær, Norcross, GA) which digitizes different signals from the instrument. Six channels in the Input/Output Module were used for two load cells (two channels), an accelerometer (one channel), a photosensor (one channel), a hydrophone (one channel), the position photosensor (one channel) and motion direction (one channel). The Pulse system does not have a separated input channel for an external trigger to start or stop the data acquisition. Instead, it has to take one analog input channel to set a condition for starting or ending the data acquisition in the software. Although this uses up one valuable channel for high speed data acquisition input, it also provides an option for automation. Epsilon Eb-205 Digital Servo Drive controls the motor moving speed, direction, and has several analog output channels which can provide voltage output. It can be programmed to output voltage synced with the motor motion direction. The voltage output is used to trigger the data acquisition. Software written in Microsoft Visual C ++ with WTL (Windows Template Library) [49] integrates the different mechanical parts of the instrument (motor, moving direction, limit switches, force for the finger) and automates data acquisition. This software controls the motor through the Emerson Epsilon Eb-205, and the power for the electromagnetic coils of “artificial finger” through the digital relays (USB-ERB24 relay) and programmable power supply (9120A, B&K Precision, Yorba Linda, CA). With the combination of the commercial software and custom-written software, different components are integrated and automation is achieved.
Fig. 3. The photographs of the apparatus. The left photo shows the whole view of the apparatus with labels for key components. The right photo shows sample stage immersed in water bath. The Points 1 and 2 are the positions for two load cells. The Points A and B are two positions to connect the posts which extend the sample stage down to the water bath. The black foam underwater is the neoprene form with Vitro-Skin on it. Point 3 is the contact position for the “finger” surface and the substrate surface. The surface wave was generated by the stick–slip events during rinse-off personal wash product.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
4
J. Ding / Biotribology xxx (2015) xxx–xxx
2.1.3. Friction coefficient The setup with two normal load cells creates a system which can be used to measure the normal load, and also derive the sliding friction and friction coefficient. Fig. 5 shows the schematic diagram of the sample stage. The diagram is an analog to the stage photo in Fig. 3. The positions 1 and 2 are the points for the two load cells (N1 and N2) which can measure the normal force. The positions A and B are the points connected to the extended posts which connect to the sample stage. Point 3 is the position where two surfaces contact each other during sliding. Since the sample substrates and the “artificial finger” are both soft materials, the contact surface might be a narrow stripe instead of a single point. In that case, Point 3 should be the middle point of the contact area to simplify the force analysis without loss of the generality. In each position, the forces in two directions are labeled such as F1f, F2f and Ff for the tangential force, Fn, N1 and N2 for the normal force. We did not consider the weight of the sample stage and the buoyant force since N1 and N2 can include both forces and will not affect our calculation. In addition, since the two load cells are not in the same vertical plane with the contact area, there should be forces toward us from the diagram in Fig. 5. It does not affect our force analysis in other directions. Here we show how to calculate frictional force and coefficient from the two normal load cell readings. Based on force balance in normal and tangential directions: F n ¼ N1 þ N2
ð1Þ
F f ¼ F1f þ F2f
ð2Þ
where Fn is the normal force applied on the substrate by the artificial finger; N1 and N2 are normal forces measured by two load cells; Ff is the friction generated by the artificial finger on the substrate; F1f and F2f are the forces in the horizontal direction at the points of two load cells. The moments at point A and point B must be balanced too. At Point A: N1 L1 þ F 1 f h1 þ F 2 f h1 þ F n x ¼ F f h þ N2 ðL þ L1 Þ
ð3Þ
At Point B: N1 ðL þ L1 Þ þ F 1 f h1 þ F 2 f h1 ¼ F f h þ F n ðL−xÞ þ N 2 L1
ð4Þ
Therefore, the friction generated by the artificial finger sliding across the substrate can be summarized by: Ff ¼
1 ððN1 þ N2 Þðx þ L1 Þ−N2 ðL þ 2L1 ÞÞ h−h1
ð7Þ
and the friction coefficient can be obtained: μ¼
1 N2 ðL þ 2L1 Þ ðx þ L1 Þ− N1 þ N2 h−h1
ð8Þ
N1, N2, and x are measurable data, and L, L1, h and h1 are fixed for the apparatus. 2.2. Materials Two commercial body wash products with very different in-use feel, Lux Soft Kiss and Lux Recarregue from Unilever, were tested. Eight personal cleanser prototypes with composition differences (different surfactants, different polymers, with or without oil, etc.) were also examined. The hardness of tap water from our lab in Trumbull, Connecticut, is usually around 55 ppm CaCO3. To adjust water hardness, CaCl2·2H2O (Mallinckrodt Baker, Inc., Phillipsburg, NJ) and MgCl2·6H2O (EMD Chemicals Inc., Gibbstown, NJ) were added to water. Both chemicals were purchased from Sigma Aldrich, MO. Water hardness was adjusted to 100 ppm CaCO3 with the atomic ratio of Ca to Mg (3:1). To mimic the softness of deep skin and firmness of the skin surface, two layers of substrates were used for both the substrates as shown in Fig. 3. The bottom layer (black layer in the photo) is a weather-resistant neoprene form (McMasterCarr, Robbinsville, NJ) with a self-adhesive to attach it to the metal sample stage and “artificial finger”. The neoprene foam is 2 in. (5.08 cm) wide, 3/8 in. thick (0.95 cm), and has a firmness of 5–9 psi (25% deflection). On the top of the neoprene form, we affixed a thin Vitro-Skin (IMS Inc., Portland, ME). The Vitro-Skin is cut to an 1″× 5″ (2.54 cm × 12.7 cm) rectangle shape and pre-treated with warm water (~ 34 °C), then bound to the neoprene foam with rubber bands. The Vitro-Skin has two sides: one with a smooth surface and another with a coarse surface. The smooth surface is used for contacting personal wash products. 3. Application examples
Rearrange Eq. (3): N1 L1 þ F n x ¼ F f h− F 1 f h1 þ F 2 f h1 þ N 2 ðL þ L1 Þ
ð5Þ
Put Eqs. (1) and (2) into Eq. (5): N1 L1 þ ðN 1 þ N2 Þx ¼ F f ðh−h1 Þ þ N2 ðL þ L1 Þ:
ð6Þ
Fig. 5. The schematic diagram for force balance. Points 1, 2, 3 and A, B are corresponding points in Fig. 3.
3.1. Wash-off procedure To prepare the test, the water bath is filled with 50 L of water adjusted to the test temperature of 34 °C and target hardness (100 ppm for the data reported in this paper), and lifted up to immerse the sample stage. The high heat capacity of 50 L of water is sufficient to keep the experimental temperature stable for the duration of testing. The initial position of the “artificial finger” is as shown in Fig. 1. The “artificial finger” is programmed to take a certain load by applying a controlled current through the magnetic coil inside it and the swiveling arm is set to a specified moving speed to slide the “artificial finger” across the substrate under water. During this period, the data acquisition system is activated to digitize loads applied to the substrates, and measure the sound pressure under water, the vibration of the “artificial finger” and the photosensor signal for the contact position. When the swiveling arm reaches the end of the first sweep (the left end reference to Fig. 3), a limit switch is depressed, the current through the magnetic coil in the “artificial finger” reverses and the “artificial finger” lifts to prevent the “artificial finger” crashing against the sample stage. The swiveling arm moves back to initial position where another limit switch is depressed. This completes one washing cycle. After 10 washing cycles baseline to condition the Vitro-Skin, the water bath is lowered to bring the substrates out of water. After the Vitro-Skin is dried with paper towel to remove excess water, 0.25 g of personal wash product
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
J. Ding / Biotribology xxx (2015) xxx–xxx
is applied to the Vitro-Skin. 0.5 mL water is applied to dilute the personal wash product. The product is spread over the surface of Vitro-Skin at minimal speed to avoid lather. Then the sample stage is re-immersed in the water bath and data is collected over a 40 washing cycle period. 3.2. Data reconstruction Data from each washing cycle are stored in a file which, as discussed above, includes data from 6 channels. Fig. 6 shows signals from a single washing cycle where stick–slip was observed. Data from 5 channels are
Fig. 6. The signals from different sensors within one washing cycle. In the figure, signal a is the gate signal from the photosensor. The rising edge corresponds to Position A in Fig. 3, the finger contact position, and the falling edge to Position B. Signal b and c are the dynamic loads from two load cells, and signal d is the sum of two loads in Signal b and c. Signals e and f are the “finger” vibration and underwater sound, respectively. The two outside dash lines are identified by the rising and falling edges in Signal a. Two inner dash lines are the border lines to extract the signals from the finger sliding across the substrate. The numbers in b are 5 stages for one washing cycle.
5
plotted since the sixth channel digitizes the motor direction signal which is used to trigger the data acquisition software and does not reflect the rinse properties. From top to bottom in Fig. 6, signals from the photosensor, load cell 1, load cell 2, total normal load, accelerometer and underwater sound are plotted, respectively. Fig. 6b to f are highly correlated since they are signals from different sensors for the same process. Fig. 6b, for example, records the whole moving finger from initial position (very right side outside the view in Fig. 3) to the end (the very left side outside of view in Fig. 3). There are 5 stages in the signal. Stage 1: the “finger” freely approaches to the sample stage and only sample stage weight is recorded by the load cells. Stage 2: the “finger” touches the edge of sample stage when the pivot changes the “finger” angle to apply the force to the substrate. Stage 3: the “finger” passes across the rubber band which is used to fast the Vitro-Skin to the neoprene foam substrate. Stage 4: the “finger” surface slides across the substrate and moves away from the load cell. In this stage, the average sensed load decreases but the total normal load average remains nearly constant (Fig. 6d). During this period, stick–slip events are clearly observed and the load jumps up and down with the “finger” vibrations. Stage 5: the “finger” slides against the substrate edge creates oscillation. The useful information from sliding surfaces is in Stage 4. Fig. 6a plots the gate signal from the photosensor; this data reduces the work to extract the useful information. The rising edge corresponds to the “finger” at position A in Fig. 5, x = 0 in Eq. (8). The falling edge corresponds to the “finger” at position B in Fig. 5, x = L in Eq. (8). Based on the two edges in Fig. 6a, the unnecessary signals for all channels can be cut out automatically; however, the signal cannot be simply cut within the gate signal because the rubber bands used to hold the substrates create vibrations which can provide false signals. The rubber band is a little higher than the Vitro-Skin surface, therefore, the “finger” needs more time to only contact Vitro-Skin surface. It is advisable to cut an additional portion of the data sequence from both ends of the gated signal to get rid of edge and rubber band effects. In practice, about 1/6 of the whole gated range from both ends is cut off (between the two dash lines in both ends in Fig. 6). After cutting out signals from both ends, a short data sequence per channel remains. Fig. 7 lists three plots from total normal load (Fig. 7a, N1 + N2), “finger vibration” in term of acceleration (Fig. 7b) and underwater sound pressure (Fig. 7c). All three are highly correlated. The total normal load shows the clean signal of the force from the jumping finger generated by the stick–slip events. The vibration of the “finger” detected by the accelerometer tracks the “finger” jumping. Due to the applied force from the magnetic coils inside the pivot of the “finger”, the jumping of the “finger” is asymmetric. The amplitude of both normal load and “finger” vibration reflects the strength of the stick–slip events. Comparing to other signals, the underwater sound is noisier due to the ripple effect from the water. The stick–slip events from two surfaces can generate observable water waves as shown in the photograph in Fig. 3. It is actually sufficient to look at just the total normal load signal to find the stick–slip events. Each washing cycle creates useful signals which last more than 0.8 s at the swiveling speed of 33 mm/s. At lower speed, one obtains longer data sequences per washing cycle, and vice versa. In the real experiment, it might take about 20 s for the artificial finger to complete one cycle at the swiveling speed of 33 mm/s. For a simple application, we take 40 washing cycles to make sure the surfactants in the personal wash products are rinsed off. It helps visually to plot the extracted signals from each washing cycle together. Fig. 8 shows an example. In each plot, the x-axis is the reconstructed timeline which is accumulated time from the first to the last washing cycles by only considering the time taken for the useful signals. For example, the first washing cycle yields the first 0.8 s of useful time, the 2nd washing cycle's 0.8 s of useful time are put just behind the signal from first washing cycle, etc. The plots in Fig. 8 illustrate the rinse profiles from two personal wash products with distinct in-use feels. Lux Recarregue (Figs. 8a and 8b) is a syndet base body wash product which provides a soft and
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
6
J. Ding / Biotribology xxx (2015) xxx–xxx
whether it has stick–slip events, how strong the stick–slip is, how long the smooth region (slimy) is, whether the stick–slip is regular or irregular, the frequency of the stick–slip, and how the friction coefficient changes during rinse, etc. Analysis of all these parameters is beyond the scope of this paper, however, we illustrate how one of these parameters, the strength of stick–slip events, defined as the standard deviation of the total normal force in each washing cycle, can be used. Consider the standard deviation of the total normal force xi in each washing cycle: vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u n u1 X ðxi −xÞ2 : σ ¼t n i¼1
ð9Þ
This quantifies the amount of variation in the total normal force sensed by the two load cells. The bigger the standard deviation is, the stronger stick–slip is. Fig. 9 plots the standard deviation of the total normal force for 40 washing cycles from 8 different personal cleanser prototypes and two commercial products, Lux Soft Kiss and Lux Recarregue. Clearly, different cleansers have different behaviors in wash-off. Sample 7, 8 and Lux Soft Kiss have very high standard deviation after the initial low standard deviation which indicates the strong stick–slip after several washing cycles although Sample 7 shows slightly less stick–slip after the initial strong stick–slip. These samples provide clean rinse feel during wash after the initial “slimy” feel. Samples 2, 4, and 5, and Lux Soft Kiss, on the other hand, show a minimal standard deviation for all 40 washing cycles. Those samples create a very slimy feel during rinse. Interesting products are Samples 1, 3, and 6 which have more washing cycles of low standard deviation and generate intermediate stick–slip. The results provide rich information of measurable parameters connecting to formulations, benefit deposition and consumer perception. Generally, the new apparatus is a powerful tool to understand the relationship between consumer perception and measurable physical parameters. 3.4. Friction coefficient Fig. 7. The extracted signals from the finger sliding the sample substrate. The “squeaky” region is shown here from one washing cycle. Signal a is the total normal load from the sum of two load cells. Signal b is the “finger” vibration detected by the accelerometer. Signal c is the underwater sound pressure detected by hydrophone.
smooth rinse feel during wash. The major surfactants in Lux Recarregue are sodium lauryl ether sulfate (SLES) and cocamidopropyl betaine (CAPB). Lux Soft Kiss (Figs. 8c and 8d), by contrast, is a soap base (blend of fatty soaps with different chain lengths) body wash which provides a clean rinse feel during wash. For Lux Recarregue, both total normal load and “finger” vibration do not show stick–slip for all 40 washing cycles. In contrast, for Lux Soft Kiss, both total normal load and “finger” vibration show two distinct regions in 40 washing cycles. The first two washing cycles do not have any stick–slip. The sliding movement of the “finger” across the substrate is very smooth; this corresponds to the slimy region. At this stage, there are enough surfactant on the skin surface (for a consumer) or on the Vitro-Skin surface (as in the apparatus reported here) to generate very slippery surfaces. Starting from the 3rd washing cycle, a strong stick–slip commences; this corresponds to the consumer's reported “squeaky-clean” feel when using his/her fingers to slide across skin surface. Generally, the apparatus results are highly correlated to in-use perceptions. 3.3. Simplification Fig. 8 provides visual plots for smooth or stick–slip movements of sliding “finger” which can be used to predict in-use perceptions for personal wash products. These plots have extremely large amount of data. Simplification of the raw data should be considered to extract meaningful parameters. There are many ways to analyze the raw data. Taking the total normal force, for example, we can analyze the data to determine:
The apparatus uses two load cells which can be used to derive friction coefficient indirectly. Fig. 10 plots the derived friction coefficients of 6 cleanser prototypes and Lux Recarregue. Lux Soft Kiss and Samples 7, 8 shown in Fig. 9 do not have meaningful friction coefficient due to the strong stick–slip during rinse. All cleansers have very low friction coefficients for the first two washing cycles because the initial local high concentration of surfactants act as lubricants. All samples except Samples 3 and 6 retain very low friction coefficients for all 40 washing cycles. The slight difference among those samples is likely not significant. Samples 3 and 6 however have significantly different trends from other samples. Sample 6 has basically the same surfactant system as Lux Recarregue with an additional hydrophobic polymer. A hydrophobic polymer with high molecular weight (typically C2–C10 polyalkenes such as polybutene here) in Sample 6 tunes the rinse and friction coefficient profile. After the surfactants on the substrate surfaces is washed off, the effect of the polymer dominates the friction coefficient during the rest of washing which results a sudden rise in friction coefficient. Sample 3 contains a surfactant combination with the major surfactant of Sodium Cocoyl Isethionate (SCI), a gentle surfactant based on coconut fatty acids. It has an intermediate friction properties. Friction coefficient changes during rinse can be used to understand the effect of different raw materials, product structuring and formulation, etc., therefore, and guide product formulation to meet consumers' expectations without incurring the much more significant costs of a consumer study. 4. Summary We reported the design and control of a custom-built apparatus to measure friction and vibration profiles during simulated washing with personal wash products. We described how to automate data collection,
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
J. Ding / Biotribology xxx (2015) xxx–xxx
7
Fig. 8. The rebuilt rinse profiles for two LUX cleansers which have different in-use feelings. The left two plots (a and c) are the total normal load, and the right two plots (b and d) are the “finger” vibration, from 40 washing cycles. The top two plots are the rinse profiles from LUX Recarregue and the bottom two plots are the rinse profiles from LUX Soft Kiss. In c and d, the stick–slip events generate the big oscillations in the signal from the 3rd washing cycle. To observe the detail of the stick–slip in each washing cycle, refer Fig. 7.
and cut unnecessary data to limit bandwidth while retaining all necessary information to reconstruct rinse profiles. We have shown that these rinse profiles differ with different cleansers in ways that have straightforward connections to the consumer tactile experience. For the soap base cleansers such as the Lux Soft Kiss product, the instrument indicates a low amplitude in the finger vibration, low noise of underwater sound, and constant normal force, during the first two washing cycles, corresponding to a very short slimy region. Following the slimy region, the soap system generates stick–slip events which trigger the strong finger vibration, higher underwater noise, and large normal
Fig. 9. The standard deviation of total normal force from different personal cleansers.
force variation. For the syndet base cleanser such as Lux Recarregue product, stick–slip events generally do not occur through the whole rinse period. The results correlate with the clean-rinse feel reported by consumers. Tuning the cleanser formulation with different techniques affects the friction coefficient during rinsing. The apparatus appears to be a promising tool to study the in-vitro rinse properties of cleansers.
Fig. 10. The friction coefficient of different personal cleansers indirectly measured from two load cells with the “artificial finger” apparatus in term of washing progress. For two samples with strong stick–slip (samples 7 and 8) shown in Fig 9, the friction efficient is not meaningful due to the discontinuous contact of two surfaces and is not shown here.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003
8
J. Ding / Biotribology xxx (2015) xxx–xxx
Acknowledgments The author acknowledges the support from and helpful discussions with Larry Wilen, KP Ananth and Albert Post. The early collaborations with Akay Adnan, Blair Echols, Anne Dussaud, Frank Meyer and Alex Lips which led to the current work are also highly appreciated. Daniel Rodriguez helped to create several drawings. The author would like to thank Heidi Warriner for proofreading of the final draft. References [1] M.H. Lee, Tactile sensing: new directions, new challenges, Int. J. Robot. Res. 19 (7) (2000) 636–643. [2] R. Dahiya, P. Mittendorfer, M. Valle, G. Cheng, V. Lumelsky, Directions toward effective utilization of tactile skin: a review, IEEE Sensors J. 13 (11) (2013) 4121–4138, http://dx.doi.org/10.1109/JSEN.2013.2279056. [3] S. Stassi, V. Cauda, G. Canavese, C. Pirri, Flexible tactile sensing based on piezoresistive composites: a review, Sensors (Switzerland) 14 (3) (2014) 5296–5332, http://dx.doi.org/10.3390/s140305296. [4] P. Shull, D. Damian, Haptic wearables as sensory replacement, sensory augmentation and trainer - a review, J. Neuroeng. Rehabil. 12 (1) (2015)http://dx.doi.org/ 10.1186/s12984-015-0055-z. [5] R. Lundström, L. Burström, Mechanical impedance of the human hand–arm system, Int. J. Ind. Ergon. 3 (3) (1989) 235–242, http://dx.doi.org/10.1016/01698141(89)90023-1. [6] H. Limerick, D. Coyle, J. Moore, The experience of agency in human-computer interactions: a review, Front. Hum. Neurosci. (2014) 8, http://dx.doi.org/10.3389/fnhum. 2014.00643 (AUG). [7] X. Liao, Y. Li, J. Hu, X. Wu, Q. Li, A simultaneous measurement method to characterize touch properties of textile materials, Fibers Polym. 15 (7) (2014) 1548–1559, http://dx.doi.org/10.1007/s12221-014-1548-2. [8] A. Klöcker, M. Wiertlewski, V. Théate, V. Hayward, J.L. Thonnard, Physical factors influencing pleasant touch during tactile exploration, PLoS ONE 8 (11) (2013)http://dx.doi.org/10.1371/journal.pone.0079085. [9] R. Samra, M. Zadeh, D. Wang, Design of a tactile instrument to measure human roughness perception in a virtual environment, IEEE Trans. Instrum. Meas. 60 (11) (2011) 3582–3591, http://dx.doi.org/10.1109/TIM.2011.2161149. [10] C. Lucarotti, C. Oddo, N. Vitiello, M. Carrozza, Synthetic and bio-artificial tactile sensing: a review, Sensors (Switzerland) 13 (2) (2013) 1435–1466, http://dx.doi.org/10. 3390/s130201435. [11] A. Hussain, A. Malik, M. Halim, A. Ali, The use of robotics in surgery: a review, Int. J. Clin. Pract. 68 (11) (2014) 1376–1382, http://dx.doi.org/10.1111/ijcp.12492. [12] H. Zahouani, S. Mezghani, R. Vargiolu, T. Hoc, M. El Mansori, Effect of roughness on vibration of human finger during a friction test, Wear 301 (1–2) (2013) 343–352, http://dx.doi.org/10.1016/j.wear.2012.11.028. [13] A. Sawers, L. Ting, Perspectives on human-human sensorimotor interactions for the design of rehabilitation robots, J. Neuroeng. Rehabil. 11 (1) (2014)http://dx.doi.org/ 10.1186/1743-0003-11-142. [14] A. Zimmerman, L. Bai, D. Ginty, The gentle touch receptors of mammalian skin, Science 346 (6212) (2014) 950–954, http://dx.doi.org/10.1126/science.1254229. [15] T. Heed, V. Buchholz, A. Engel, B. Röder, Tactile remapping: from coordinate transformation to integration in sensorimotor processing, Trends Cogn. Sci. 19 (5) (2015) 251–258, http://dx.doi.org/10.1016/j.tics.2015.03.001. [16] B.N.J. Persson, Sliding Friction physical Principles and Applications, NanoScience and Technology, Springer-Verlag, Berlin Heidelberg, 2000http://dx.doi.org/10.1007/9783-662-04283-0. [17] M. Adams, B. Briscoe, S. Johnson, Friction and lubrication of human skin, Tribol. Lett. 26 (3) (2007) 239–253, http://dx.doi.org/10.1007/s11249-007-9206-0. [18] M. Adams, S. Johnson, V. Lefèvre, P. Lévesque, V. Hayward, T. André, J.L. Thonnard, Finger pad friction and its role in grip and touch, J. R. Soc. Interface 10 (80) (2013)http://dx.doi.org/10.1098/rsif.2012.0467. [19] S. Pasumarty, S. Johnson, S. Watson, M. Adams, Friction of the human finger pad: influence of moisture, occlusion and velocity, Tribol. Lett. 44 (2) (2011) 117–137, http://dx.doi.org/10.1007/s11249-011-9828-0. [20] S. Schreiner, M. Rechberger, J. Bertling, Haptic perception of friction — correlating friction measurements of skin against polymer surfaces with subjective evaluations of the surfaces' grip, Tribol. Int. 63 (2013) 21–28, http://dx.doi.org/10.1016/j. triboint.2012.08.017. [21] A. Takahashi, M. Suzuki, Y. Imai, Y. Nonomura, Tactile texture and friction of soft sponge surfaces, Colloids Surf. B: Biointerfaces 130 (2015) 10–15, http://dx.doi. org/10.1016/j.colsurfb.2015.03.055. [22] A. Gallace, C. Spence, The cognitive and neural correlates of “tactile consciousness”: a multisensory perspective, Conscious. Cogn. 17 (1) (2008) 370–407, http://dx.doi. org/10.1016/j.concog.2007.01.005.
[23] T. Whitaker, C. Simões Franklin, F. Newell, Vision and touch: independent or integrated systems for the perception of texture? Brain Res. 1242 (2008) 59–72, http://dx.doi.org/10.1016/j.brainres.2008.05.037. [24] Y.-C. Pei, T.-Y. Chang, T.-C. Lee, S. Saha, H.-Y. Lai, M. Gomez-Ramirez, et al., Crossmodal sensory integration of visual–tactile motion information: Instrument design and human psychophysics, Sensors (Switzerland) 13 (6) (2013) 7212–7223, http://dx.doi.org/10.3390/s130607212. [25] S. Soto-Faraco, G. Deco, Multisensory contributions to the perception of vibrotactile events, Behav. Brain Res. 196 (2) (2009) 145–154, http://dx.doi.org/10.1016/j.bbr. 2008.09.018. [26] V. Occelli, C. Spence, M. Zampini, Audiotactile interactions in front and rear space, Neurosci. Biobehav. Rev. 35 (3) (2011) 589–598, http://dx.doi.org/10.1016/j. neubiorev.2010.07.004. [27] G. Deco, R. Romo, The role of fluctuations in perception, Trends Neurosci. 31 (11) (2008) 591–598, http://dx.doi.org/10.1016/j.tins.2008.08.007. [28] J. Foxe, Multisensory integration: frequency tuning of audio-tactile integration, Curr. Biol. 19 (9) (2009) R373–R375, http://dx.doi.org/10.1016/j.cub.2009.03.029. [29] J. Löfvenberg, R. Johansson, Regional differences and interindividual variability in sensitivity to vibration in the glabrous skin of the human hand, Brain Res. 301 (1) (1984) 65–72, http://dx.doi.org/10.1016/0006-8993(84)90403-7. [30] H. Hämäläinen, T. Järvilehto, K. Soininen, Vibrotactile atonal interval correlated with activity in peripheral mechanoreceptive units innervating the human hand, Brain Res. 333 (2) (1985) 311–324, http://dx.doi.org/10.1016/0006-8993(85)91585-9. [31] G. Gescheider, S. Bolanowski Jr., R. Verrillo, Vibrotactile masking: effects of stimulus onset asynchrony and stimulus frequency, J. Acoust. Soc. Am. 85 (5) (1989) 2059–2064. [32] G. Gescheider, S. Bolanowski Jr., R. Verrillo, D. Arpajian, T. Ryan, Vibrotactile intensity discrimination measured by three methods, J. Acoust. Soc. Am. 87 (1) (1990) 330–338. [33] J. Bell, S. Bolanowski, M. Holmes, The structure and function of pacinian corpuscles: a review, Prog. Neurobiol. 42 (1) (1994) 79–128, http://dx.doi.org/10.1016/03010082(94)90022-1. [34] M. Hollins, E. Roy, Perceived intensity of vibrotactile stimuli: The role of mechanoreceptive channels, Somatosens. Mot. Res. 13 (3–4) (1996) 273–286. [35] A. Brisben, S. Hsiao, K. Johnson, Detection of vibration transmitted through an object grasped in the hand, J. Neurophysiol. 81 (4) (1999) 1548–1558. [36] K. Johnson, T. Yoshioka, F. Vega Bermudez, Tactile functions of mechanoreceptive afferents innervating the hand, J. Clin. Neurophysiol. 17 (6) (2000) 539–558. [37] G. Gescheider, S. Bolanowski, K. Hardick, The frequency selectivity of informationprocessing channels in the tactile sensory system, Somatosens. Mot. Res. 18 (3) (2001) 191–201, http://dx.doi.org/10.1080/01421590120072187. [38] K. Johnson, The roles and functions of cutaneous mechanoreceptors, Curr. Opin. Neurobiol. 11 (4) (2001) 455–461, http://dx.doi.org/10.1016/S09594388(00)00234-8. [39] G. Gescheider, S. Bolanowski, J. Pope, R. Verrillo, A four-channel analysis of the tactile sensitivity of the fingertip: frequency selectivity, spatial summation, and temporal summation, Somatosens. Mot. Res. 19 (2) (2002) 114–124, http://dx.doi.org/10. 1080/08990220220131505. [40] G. Gescheider, S. Bolanowski, R. Verrillo, Some characteristics of tactile channels, Behav. Brain Res. 148 (1–2) (2004) 35–40, http://dx.doi.org/10.1016/S01664328(03)00177-3. [41] B. Güçlü, S. Bolanowski, Vibrotactile thresholds of the non-pacinian i channel: I. methodological issues, Somatosens. Mot. Res. 22 (1–2) (2005) 49–56, http://dx. doi.org/10.1080/08990220500084719. [42] B. Güçlü, S. Bolanowski, Vibrotactile thresholds of the non-pacinian i channel: II. predicting the effects of contactor location on the phalanx, Somatosens. Mot. Res. 22 (1–2) (2005) 57–68, http://dx.doi.org/10.1080/08990220512331387971. [43] A. Akay, Acoustics of friction, J. Acoust. Soc. Am. 111 (4) (2002) 1525–1548, http:// dx.doi.org/10.1121/1.1456514. [44] R. Fagiani, F. Massi, E. Chatelet, Y. Berthier, A. Akay, Tactile perception by friction induced vibrations, Tribol. Int. 44 (10) (2011) 1100–1110, http://dx.doi.org/10.1016/j. triboint.2011.03.019. [45] R. Fagiani, F. Massi, E. Chatelet, J. Costes, Y. Berthier, Contact of a finger on rigid surfaces and textiles: friction coefficient and induced vibrations, Tribol. Lett. 48 (2) (2012) 145–158, http://dx.doi.org/10.1007/s11249-012-0010-0. [46] J. Ding, A. Dussaud, A. Lips, A novel method for characterizing tactile perception, Abstracts of Papers, 227th ACS National Meeting, Anaheim, CA, United States, March 28-April 1 2004, p. 2004 COLL-420. [47] A. Dussaud, J. Ding, A. Lips, Method and system for characterizing tactile perception, U.S. Pat. Appl. Publ. (2005) 20050106103 A1 (US 20050106103 A1). [48] A. Akay, B. Echols, J. Ding, A. Dussaud, A. Lips, Measurement of sound, vibration and friction between soft materials under light loads, Wear 276–277 (2012) 61–69, http://dx.doi.org/10.1016/j.wear.2011.12.004. [49] M. Dunn, WTL for MFC programmers, 2005http://www.codeproject.com/Articles/ 3867/WTL-for-MFC-Programmers-Part-II-WTL-GUI-Base-Class.
Please cite this article as: J. Ding, An apparatus for sound, vibration and friction measurements of soft materials in aqueous environments, (2015), http://dx.doi.org/10.1016/j.biotri.2015.11.003