夹爪抓东西一抓就滑掉，手上没点摩擦力。触觉皮肤，先把防滑耐磨做出来。
一抓就掉，不是夹不紧是摩擦不够
机器人触觉皮肤是“接触件”：防滑、耐磨、触感。材料要扛住接触——结论先给：夹爪接触面，防滑耐磨损——防滑，是触觉皮肤的底线。
机器人触觉皮肤最大的坑：夹爪打滑磨花，触觉反馈直接失灵——防滑耐磨，是触觉皮肤的两条腿。
机器人件是体验件：手感、耐用都是问题。材料选对，体验才稳——机器人件，别省配方。
防滑耐磨都要，TPE为什么合适
机器人触觉皮肤用 TPE 的理由：防滑可做、耐磨可做、触感可做——三条合起来，适合触觉。
防滑是底线：夹爪不打滑。防滑测试写进验收——打滑，就是问题。
耐磨不能省：接触面不磨花。耐磨测试写进验收——磨花，就是问题。
防滑、耐磨、触感，三关各看什么
防滑工况：抓取不打滑。防滑数据要验——打滑，就是问题。
耐磨工况：反复接触。耐磨数据要验——磨花，就是问题。
触感工况：软硬适度。手感数据要验——偏硬，就是问题。
触觉皮肤材料，一张表对照清楚
| 材料 | 防滑 | 耐磨 | 触感 |
|---|
| TPE | 好 | 好 | 好 |
| 硅胶 | 好 | 中 | 好 |
| TPU | 中 | 强 | 中 |
| 橡胶 | 好 | 好 | 好 |
表格读法：材料按任务排，任务不同选不同，摩擦系数和磨耗都要看。
防滑耐磨一起看，材料才选对。
抓得稳磨不穿，验收测这四项
| 项目 | 要求 | 判断 |
|---|
| 防滑 | 测试 | 达标 |
| 耐磨 | 测试 | 达标 |
| 触感 | 盲测 | 达标 |
| 批次 | 锁定 | 达标 |
表格读法：触觉皮肤一项项核，体验看得见，磨耗按万次循环测。
防滑，是触觉皮肤的底线。
只追触感不测防滑，抓什么都掉
坑一：只看触感。触感细腻却不防滑，夹爪一抓就溜——防滑必测。
坑二：耐磨漏测。磨花——耐磨必测。
坑三：批次不锁。手感漂——批次锁定。
摩擦、耐磨、触感——做皮前问清
三问：什么抓取物、什么耐磨次数、什么触感。一验：实际使用实测——三问一验，供应商底细清楚。
防滑验证要先行：先把动摩擦系数和万次磨耗测出来，再谈触感——防滑，是触觉皮肤的底线。
留样要成习惯：每批留样，防滑耐磨按批次复测。批次换料先对比再放量——批次稳，客诉少。
打滑、磨穿、发粘：故障对照一张表
| 现象 | 原因 | 对策 |
|---|
| 打滑 | 防滑差 | 换高防滑料 |
| 磨花 | 耐磨差 | 换耐磨料 |
| 偏硬 | 硬度高 | 调配方 |
| 脱层 | 基材不净 | 清洁基材 |
| 批次漂 | 配方波动 | 锁窗口 |
夹爪一捏就打滑掉件，客户说手感不对——防滑不到位，抓握就是空谈。 抓不住东西，再好看的触觉皮肤也白搭。
防滑好不代表耐磨，硅胶防滑好但耐磨中等，天天夹取的夹爪要耐磨加防滑一起。 两个指标同时达标才算合格。
夹爪接触面摩擦系数要按夹持物料定，光滑件和粗糙件两回事。 用真实夹取物做摩擦测试，别用钢板数据套软物。
触觉皮肤是夹爪的“指腹”，防滑耐磨触感三位一体。 触感盲测加耐磨磨耗，两项达标才放行，别只摸一下就过。
磨耗≤150mg、夹持2万次不滑脱，夹爪触觉皮肤良率稳定走量。 别只看新件手感，要看用过后还滑不滑。
触觉皮肤摩擦系数要0.6以上，用真实夹取物测别用钢板。钢板摩擦数据套软物，抓海绵抓纸张全打滑，表面纹路深度0.3-0.5mm防滑才出得来。
触觉皮肤耐磨按Taber CS-10磨耗<50mg/1000转验收。夹爪天天开合接触面磨损快，SEBS基加耐磨剂比纯SEBS寿命长两倍，两项达标才放行。
触感盲测加耐磨磨耗两项达标才放行，别只摸一下就过。用真实夹取物做摩擦测试别用钢板数据套软物，SEBS基加耐磨剂寿命长两倍。
抓不住东西再好看的触觉皮肤也白搭。两个指标同时达标才算合格，用真实夹取物做摩擦测试别用钢板数据套软物。
科隆客户案例：耐油不足溶胀变形，调参数过老化关
扬州一家机器人零部件厂，触觉皮肤耐油性不足，接触油后溶胀变形。科隆配合调整注塑参数（模温/料温/保压），溶胀消除，一次性通过 1000h 老化测试。参数窗口锁死，溶胀从源头断——耐油问题，先看配方再看参数。
小结
机器人触觉皮肤的选型，防滑先测，耐磨再核，夹爪接触面防滑耐磨损，防滑是底线。
我们交付的，不只是一包料。
The gripper slips as soon as it grabs something; there’s no friction on the hand. For tactile skin, first make it anti-slip and wear-resistant.
It falls off as soon as you grab it; it's not that the grip isn't tight enough, it's that there's not enough friction.
Robot tactile skin is a 'contact component': anti-slip, wear-resistant, and tactile. The material must withstand contact — here’s the conclusion first: the contact surface of the gripper should be anti-slip and wear-resistant — anti-slip is the baseline for tactile skin.
The biggest pitfall of robot tactile skin: the gripper slips and gets scratched, causing tactile feedback to fail directly—anti-slip and wear-resistant properties are the two legs of tactile skin.
Robot parts are experiential parts: feel and durability are issues. Only by choosing the right materials can the experience be stable—don’t skimp on the formula for robot parts.
You want both slip resistance and wear resistance, why is TPE suitable?
Reasons for using TPE for robot tactile skin: it can be made non-slip, it can be made wear-resistant, it can be made to feel good to the touch — put together, these three make it suitable for tactile sensing.
Anti-slip is the bottom line: the gripper must not slip. Anti-slip testing is written into the acceptance—if it slips, it's a problem.
Wear resistance cannot be compromised: the contact surface should not get scratched. Include wear resistance testing in the acceptance criteria — scratches indicate a problem.
What to look at for slip resistance, wear resistance, and touch: three aspects
Anti-slip condition: Grab without slipping. Anti-slip data must be verified — slippage is a problem.
Wear-resistant condition: repeated contact. Wear-resistant data must be verified—if it's scuffed, that's a problem.
Tactile condition: moderately soft and hard. Hand-feel data must be checked—if it's on the hard side, that's a problem.
Tactile skin materials, clearly compared in one table
| Material | Non-slip | Wear-resistant | Touch |
|---|
| TPE | Good | Good | Good |
| Silicone | Good | middle | Good |
| TPU | middle | Strong | middle |
| Rubber | Good | Good | Good |
How to read the table: Materials are arranged according to tasks, choose differently for different tasks, and you need to look at both the friction coefficient and wear.
Look at both slip resistance and wear resistance together to choose the right material.
Grasp firmly but cannot grind through, test these four items during inspection
| Project | Requirement | Judgment |
|---|
| Non-slip | Test | Meet the standard |
| Wear-resistant | Test | Meet the standard |
| Touch | Blind test | Meet the standard |
| Batch | Lock | Meet the standard |
Table reading method: Each tactile skin item is checked, the experience is visible, and wear is measured by ten thousand cycles.
Anti-slip is the baseline of tactile skin.
Only chasing the feel without considering slip resistance, everything slips out of your grip.
Pitfall 1: Only looking at the texture. A texture may be delicate but not non-slip; the claw slips as soon as you grab it — slip resistance must be tested.
Pitfall 2: Abrasion resistance not tested. Scratches—abrasion resistance must be tested.
Pitfall 3: Batches are not locked. The feel is inconsistent — lock the batches.
Friction, wear resistance, touch—ask clearly before making leather
Three questions: what is being gripped, the number of wear-resistant uses, what the touch feels like. One test: actual use measurement—three questions and one test make the supplier’s details clear.
Slip resistance verification must come first: first measure the dynamic friction coefficient and ten-thousand-cycle wear, then talk about the tactile feeling—slip resistance is the baseline of tactile skin.
Making sample retention a habit: retain samples for each batch, and retest slip resistance and wear resistance batch by batch. When changing materials between batches, compare first before increasing the volume—stable batches result in fewer customer complaints.
Slipping, wearing out, sticky: a table for troubleshooting
| Phenomenon | Reason | Countermeasure |
|---|
| slip | Poor slip resistance | Replace with high anti-slip material |
| frosted | Poor wear resistance | Replace wear-resistant material |
| Somewhat hard | High hardness | Formulate a prescription |
| Delamination | The base material is not clean | Clean the substrate |
| Batch bleaching | Formula fluctuation | Lock window |
The gripper slips and drops parts with just a pinch; the customer says the feel is wrong—anti-slip is inadequate, and the grip is just empty talk. If it can't hold anything, even the most attractive tactile skin is useless.
Good slip resistance does not mean good wear resistance. Silicone has good slip resistance but medium wear resistance. Grippers that are used every day need both wear resistance and slip resistance. Both indicators must meet the standards to be considered qualified.
The friction coefficient of the gripper contact surface should be determined according to the material being clamped; smooth parts and rough parts are two different matters. Perform friction tests with the actual objects to be gripped, and don’t use steel plate data for soft objects.
The tactile skin is the 'finger pad' of the gripper, combining anti-slip, wear resistance, and tactile sensation in one. The tactile blind test plus wear resistance test must both pass before approval; don’t just touch it once and consider it okay.
Wear ≤150mg, clamping 20,000 times without slipping, the tactile skin on the clamps maintains stable yield and volume. Don't just look at the feel of new parts; you need to see if they still slip after use.
The tactile skin friction coefficient should be above 0.6. When testing with real objects, do not use steel plates. Using steel plate friction data for soft objects will result in slipping when grabbing sponges or paper. A surface texture depth of 0.3-0.5mm is needed for anti-slip performance to be effective.
Tactile skin abrasion resistance test with Taber CS-10 shows wear <50mg/1000 revolutions for acceptance. The contact surface of the claw opens and closes every day and wears quickly; SEBS-based material with added wear-resistant agent lasts twice as long as pure SEBS. Both criteria must be met for release.
Only release after passing both tactile blind tests and wear abrasion standards; don’t just touch it once and pass. Use real clamped objects for friction testing, don’t apply steel plate data to soft materials. SEBS-based material with added wear-resistant agents lasts twice as long.
No matter how good the tactile skin looks, it’s useless if it can’t grasp things. It’s only considered qualified if both indicators meet the standard. Use real objects for friction testing instead of using steel plate data on soft materials.
Cologne client case: insufficient oil resistance causing swelling and deformation, parameter adjustment leading to over-aging issues
A robot parts factory in Yangzhou found that their tactile skins lacked oil resistance, swelling and deforming when in contact with oil. Cologne helped adjust the injection molding parameters (mold temperature/material temperature/holding pressure), eliminating the swelling, and the parts passed the 1000-hour aging test on the first attempt. The parameter window was locked, stopping swelling at the source—the oil resistance issue should first examine the formulation, and then the parameters.
Summary
Selection of tactile skin for robots: test for slip resistance first, then check for wear resistance. The contact surface of the gripper should be slip-resistant and wear-resistant, with slip resistance being the bottom line.
What we deliver is not just a package of materials.
