Seeing and Solving Tribology Problems

Interview conducted by Stuart MilneApr 9 2018

Dr Don Cohen, Principal of Michigan Metrology LLC, talks about how Michigan Metrology has been at the heart of solving some of the most perplexing issues in the US auto industry. In this interview he gives an informed perception about what it takes to identify and solve problems related to surfaces, tribology and friction.

How did you get into tribology?

I grew up in Michigan in the 60s and 70s and at that time the auto industry was constantly in decline. I remember people literally vandalizing Japanese cars because they felt they were taking jobs away. I was an engineer, a technologist, a physicist. I wanted to head out to Silicon Valley, so I left.

I moved to Arizona and did graduate work in physics and optics. I ended up working with IBM and was in San Jose a lot. I worked with Wyko who made optical profilers. Wyko was getting involved in the field of tribology. My introduction in tribology came, actually, from the disk drive industry. I met some of the bigger players in that industry through IBM, and that began my fascination with tribology. Tribology is the study of friction, wear and lubrication. The disk drive industry, at least the old disk drives, actually had little heads flying over disks and that was a tribology issue. At Wyko, we developed instruments to measure what we called rougher parts because we were getting an interest in what we could offer the automotive industry.

Tribology: Seeing and Solving Problems from AZoNetwork on Vimeo.

Some of our first customers from the newer technology we developed were from the automotive industry. They wanted to measure things like pistons and crankshafts and clutch plates and we didn't even know what these things were. We measured silicon wafers and intraocular lenses and stuff like that.

Anyway, it rekindled my interest in Detroit. When I was with Wyko we actually did our first demo in Detroit, a place called Sterling Heights. I remember flying in from San Francisco and landing in the Detroit airport, looking at the map and saying, "I grew up here, but I don't really know where this place is." I realized I had grown up in my little square mile of Michigan, went to school and left. Where we had to go was a place called Sterling Heights where Ford has a huge factory. We did our first demo measuring clutch plates and I realized just how amazing the auto industry is from a tribology point of view. Tribology, the field of friction, wear and lubrication, it’s all about cars, particularly cars of this era. As we move into electric cars and new technology, the degree of tribology may decrease.

I ended up moving back to Detroit. And I said to myself, “why not take a chance and set up a business here where I just simply do tribology measurements for people using the optical profiler from Wyko?” Wyko is now part of Bruker. So, I used those optical profilers to develop applications in the automotive industry.

How different are the problems in electronics compared to the automotive industry?

There are more engineers per capita in the Detroit area than Silicon Valley. It's amazing how much technology is here because a car requires so much technology, starting from the basic of a piston and piston ring sliding up and down the cylinder bore, and going all the way out to the steering wheel. The technology to get there is enormous. I dug into the technology and I asked, “what drives this industry?” The disk drive industry in Silicon Valley was all about how fast can you go and how much data storage can you handle. What was surprising to me was that in the automotive industry, it wasn't the high-tech things we like to get interested in, such as torque and horsepower. I mean those are important things, but there are three things that really drive the buyer and therefore the industry.

Believe it or not number one is styling. I say believe it or not because I really don't care how my car looks as long as it gets me from point A to point B. For the average buyer or consumer though, it's all about styling. How does that car look to them? How does the paint look to them? How does it feel to them physically when they sit in their car? That whole concept of styling is where some of the biggest investments are made. I've seen projects over the years that blow your mind with the kind of intensity and the high-level people that you have involved, super high-level PhD scientists trying to solve paint issues and things like that.

Image Credits: Ayman alakhras/shutterstock.com

Number two is safety. Is the car going to be safe, because after all, you don't want to get killed in it. Now, when you look at what's going on, it's a very exciting time because we're finally going to use electronics to avoid accidents. I mean even if we never get to autonomous driving, the idea that my car can be driving down the road and see a collision about to happen ahead of me and shut both the other car and my car off so we don't crash, well, it's just enormous. And that's being developed, and that will happen. Safety is a big issue. The kinds of things we do to improve the safety are amazing. From seatbelts all the way to airbags, the technology that's involved is enormous. When you think about how you make an airbag work, how you make something deploy so quickly that you don't crash and you don't blow up in the process. It takes an enormous amount of technology, from the little sensors to the fabrics.

Number three is what I call warranty, but it's actually reliability. I don't want my car to break down. If my car breaks down, I want to be sure I'm safe and I can get back. We have in our cars now a mode called the Limp Home Mode. I don't know if you guys know about that, but if there is a car failure, your car hopefully will not stop. You should be able to limp on home, get to a safe place, get to a dry place where they can fix the car. It's happening too in my car, some sensor goes or something and all of a sudden, my car won't go any faster than 30 miles an hour. But that's good enough. You can get off the road. You can get to a gas station. That kind of technology is here.

What’s a common problem that you have worked on and solved?

A good problem that everybody can relate to is sun visors. When you think about it, it's a really classic tribology problem because it has almost all the aspects: friction, wear and lubrication. Basically, when you take a sun visor, your goal is to try and position that visor when you need it at a certain position, and have it stay there. You need it to move smoothly. You don't want it to squeak when it moves, and you want it to stay where you leave it. If it doesn't stay where you leave it and it falls, then it actually becomes a safety issue. There are safety issues, as well as pleasability and customer satisfaction. The industry takes sun visors very seriously here. When you look at the mechanism, you have a rod and a tube rotating around it, and the friction is key, but you don't want it to wear out. That's the wear issue. You want to maintain a certain degree of friction, so it stays in place. If there's too much friction, it would be very hard to move the thing. If there's too little friction, of course it will fall. There's quite a lot that goes on in the design of that mechanism. They treat the surfaces of those rod elements by shock blasting them to create different surface textures and also, they might coat them with different coatings. Sometimes they apply lubricants although they try and stay away from these because they will degrade.

Tribology is everywhere. What are the kinds of problems that we can fix by knowing about tribology and adjusting tribology?

A good example of one that I worked on many years ago is a highlight of some of my early years here in Detroit, and it demonstrates just how seriously the auto industry takes things like noise. Brake noise in particular. This happened to be a serious problem. This was a situation where a brand-new car would come off the assembly line. They'd step on the brakes and the brakes would make this huge noise. The interesting thing was, if you stepped on the brakes maybe 30, 40 times, it would eventually go away. But if you were to buy a $50,000 car, and they said, "Don't worry about the brakes, just step on the brakes 30 times, the noise will go away," it's not going to happen. They don't ship cars that way.

Bruker's UMT TriboLab Modules

This was a pretty serious event where they were literally testing the car off the plant floor. If it made the noise, they would drive the car into a lot until they could fix the brakes. Find out what's wrong with it, and so the lot started to fill up with cars making noise. It gets pretty scary when that happens. It turns out this was a great tribology problem to work on because it had to deal with friction. It wasn't because the surfaces were smooth, they actually were pretty rough for a brake surface. Unfortunately, they were rough in such a way that it created noise.

What was going on was that every few hundred rotors made, they changed the tool set-up. Four or five hundred made per tool setup. Every once in a while, some rotors would come off the line that would cause this noise problem. They didn't quite understand what it was. We looked at the surface texture as best we could and after a while, we got smart about it and we said, “The minute it starts to make that noise, stop the car and get me the brake rotors." Don't drive it anymore because after 10, 20 brake pushes it would go away. We were working around the clock to solve the problem. We looked at a brake that wasn't making the noise. We looked at the brake rotor that was making the noise, and there was a clear difference in the surface texture. They all met the surface texture specification at the plant. They were well within the specification. One happened to have a different looking surface texture. As a result, the rougher one was grabbing the brake pad and driving that brake pad across the face of the rotor. The noise was the brake pad hitting the end of its top, clicking, and then breaking free and continually cycling.

Once we identified that, then it was a question of how to fix the problem. This is where sometimes in tribology, things can go bad in a hurry. The people making the rotors were saying, “Well, we're meeting the spec, what can we do differently?” So they said, “Well, here's what we’ll do. Let's not wait until 500 rotors to change the tools. Let's change the tools every 200 rotors.” The minute they changed the tools from 500 rotors to 200 rotors, the problem got worse. It turned out it was the early parts in the production process that actually had the problem, and once the tools wore in and they started manufacturing the brake rotors, the later ones worked fine. The original tools, the nice clean sharp tools were cutting these very, very nice texture patterns that looked like teeth, and those teeth were biting the brake pad and driving it across the surface. It was a classic example of assuming an old tool is a bad tool. Actually, it was the good tool that was causing the problem, but surprise-surprise in the end it wasn't really the tool at all.

That was the real challenge in this problem, because it occurred seasonally. It seemed to come and go. It would happen in February, March, and April, and then go away. Finally, a colleague solved the problem. He said, “This year, we're going to solve it because we've got the two smoking guns,” and he worked very hard on it. He found that the problem had nothing to do with the tools, because they would work for the rest of the year. It was actually all in the metallurgy. What was going on was a result of weather conditions, the temperature and humidity at the foundries, were causing slight changes in the metallurgy of the cast iron. As a result, since the metallurgy was different, the machinability was different. The surface texture pattern that was being created was different, so the problem would come and go with the temperatures, and so forth, and it was quite a challenge.

The solution was to make the machining operation more robust to the metallurgy changes. That was done both with some tooling changes but also by some additional processes that were put in to make sure these really sharp teeth on the rotors weren't created during the manufacturing process. That's the one that started my business in 1997.

Tribology is everywhere, what does this mean and why is it so important?

The gecko lizard is a good example of this. It can climb up what looks like perfectly smooth glass. How can it do that? How does it get this frictional force? Let's remember, it's able to climb, so it actually can stick and release over and over again. It’s one thing to make an adhesive that would stick to the glass, but how about one that sticks and then releases, allowing you to move up?

Image Credits: sarayuth3390/shutterstock.com

It turns out that surface texture plays a part of it, but it is also little microvilli. These little “fingers” come out to provide a strong amount of adhesion between those fibers and the glass. It isn't the actual strength of the bond that does it. What makes it work is the length of those fibers because they're making a very long bond throughout that whole surface texture, even though each little bond is actually small in terms of strength. That's what allows them to move along like a caterpillar. They only have to lift off one bond at a time and since the individual bonds are not strong, they just move it along, one at a time like a ripple in a rug. Each bond is very weak, but there are a lot of them. They're able to open up one bond at a time and just walk along the glass.

Another example would be using a pencil. This is classic tribology because how do you get a pencil lead to form a line? You're basically using wear. That's a controlled amount of wear, and the graphite wears off just the right amount. Since it's graphite, it has its own lubricating property, so it writes smoothly. That's just one classic example of tribology in day-to-day life.

In the medical field, a lot of friction issues come up. In the last 20 years the amount of what we call non-invasive surgery has been increasing. So, instead of opening somebody up to get inside their heart, they actually feed catheters through your arteries. That’s a big challenge from a tribology point of view. Usually, some kind of metal material like a wire is used with a very smooth surface finish. They feed that wire through your arterial system. It's not like the arterial system is one straight tube, it's a very circuitous path. This wire has to go in all kinds of weird directions and not pierce any artery. Then they get that wire through your body into your heart, and that's just a guide wire. Now, they slide a conduit over that wire. All that sliding, that's serious tribology. You want to make sure that there is as little friction as possible both on the metal guide wire on the inside, the lumen of the tube, and on the exterior, which is sitting on the blood vessel wall. The actual physics that goes in to the development of those surfaces, their surface texture as well as the chemistry of the material is phenomenal. The friction coefficients quoted are ultra-low. They feed little tools and instruments up through the conduit, and they're able to do surgery on you without massively opening up your body. It's all about tribology and being able to feed the wire and tube through.

What about tribology and lubrication. Have you got an example of where they interact?

Probably one of the more obvious ones is skis. That's a huge tribology field, and there are all kinds of tricks that people use on skis to reduce the friction. But the one I read recently ties in with the lubrication aspect. Basically, lubrication came about as a solution to all the friction and wear problems that were happening in 1800s. Once man discovered mineral oil, we stopped worrying about friction and we just worried about keeping oil in the interface. That's basically what's happened. A lot of lubrication issues relate to the surface texture and the ability for the surface texture to hold the lubricant and to control where it flows. Often people will make parts that'll work fine. But when they duplicate the process of manufacturing the part, the part starts to fail. Sometimes it turns out one machine turns the part in a clockwise direction, but with the newer machine, they're turning the part in a counter-clockwise direction. As a result, the lubricant is pumped in one side and pumped out in the other. Simple little problems like that determine whether they keep the lubricant in.

Image Credits: Videoblocks

A nice little example is penguins and the way they swim. How do you lower the friction through water? Penguins are interesting. Some research shows that they're actually swimming on an air-bearing. Well, one of the better lubricating systems is actually to use air as a lubricant. This is what's done in the disk drive industry, the heads fly over the disk and use a little cushion of air. Similarly, the penguins come up to the surface, capture air in their feathers, and then when they dive down they're using that air layer as a little air-bearing.

How do you determine if you can help a client?

A lot of my work is generated through referral. The problem could be something like my brakes are making noises. The first is what I call the scale of interaction. Do you know where the noise is coming from? Is it indeed the brake pads engaging on the brake, or is there vibration from somewhere else? Because if it's something vibrating, it's probably not related to my work, it might be a structural problem instead. If the noise is emanating from two parts coming together, then I ask, “What's the size of the parts that are coming together? Can I even measure those parts?" Or "Can we cut a piece of the part away?" Or do we have to make what we call a replica of the part, where we actually make an impression of the part and measure it?

How do you know what size you can handle on your Bruker instruments?

A typical instrument can handle a sample about 10 inches by 10 inches by 3 inches. We can usually measure it pretty well ourselves as long as it's not too heavy, maybe below 30 pounds. Usually it comes down to the interface between a few parts and we can get to the key parts, and they may not be that big. A brake rotor on a typical car is pretty much the biggest thing I want to deal with. Even truck rotors like an F150 truck rotor can get pretty big, and it's not that easy to deal with. For the most part, the instrumentation we have from Bruker can handle pretty much any material. It could be highly reflecting. It could be dull finished. I can measure black rubber surfaces and get a really good measurement.

Most materials we can measure. The only other thing to watch out for is if the material is composed of multi-layers of films that are transparent. That can cause some problems when we're using an optical instrument. Other than that, as long as it can fit under the microscope, we can measure it. Once I can measure it, then it's a question of understanding what the potential mechanism is, if it's a noise issue, if it's a wear issue, if it's a leakage issue, if it's an appearance issue etc. What is the actual issue? How do they describe it?

If somebody has a part that is too big, what do you tell them?

Actually, I just had a problem like that. The part itself, which was part of a transmission, weighed 150 pounds and was massive. There's no way I could put that on my microscope but I could, on this particular bearing surface, make a replica of the surface. There's a material called Reprorubber. It's just a two-part material you squirt on the surface. It sets on the surface. You wait about 10-15 minutes. You peel it off and it makes a replica of the surface. However, I would much rather measure the original surface just to eliminate that one variable. In this case, we did a few projects with replicas, and the results were quite interesting. In the end we cut a section from the sample to put under the microscope.

Is there such a thing as a routine measurement? Can you give an example?

There are routine measurement requirements, so I'll give you an example. This happens in the medical industry quite a bit. This is a particular application where it's an electrode that's used in an electrotherapy device. It's implanted in somebody's body. It might be a heart implant, but it's an electrode. The electrode has to have a very particular surface texture on it so it becomes integral to the muscle tissue. For the first phase of this project, we experimented with different surface textures until the customer agreed to a specification. Then they went into production. Then routinely, once a month, I'll get a series of parts to measure before they release a lot into production. We've kept this history of measurement and data analysis for a number of years, so they can tell whether they're drifting or not in their production, and correlate that to the performance of the device.

You also get recalls in the implant industry. A good example might be a hip implant. All of a sudden, the hip implant material is not bonding well to the bone and it becomes loose, and that's not a good thing. Mostly this relates to the basic surface roughness of the part, and there's a spec on the surface texture. It is usually obvious that one is not the right texture. A company will go through the development process and they will determine a measurement protocol for that material, and that gets published with the FDA when they get their approval for their device.

What are some of the craziest, strangest, most interesting things you've measured?

The one that always takes the cake is turkey skin. That was the greatest story. Once again, surface texture, tribology comes up in a lot of places. Think of the tribology of eating. When we bite into something, what we're experiencing when we consider “this tastes good, this is crunchy,” that's all about tribology. That's all about the friction and the wear, the wear being the wear of the parts we're eating, breaking down, and the friction being that sensation we have in our mouth, that's tribology. It turns out this person we were working with was involved in poultry science. They were breeding turkeys for food, and the texture of the skin affected the way we perceive the taste of that turkey. Again, I thought that was kind of a bizarre thing, but we measured it.

Another example along the same lines: A lady from the USDA called me with an idea. Her theory was that perhaps you can detect whether eggs are tainted by looking at the surface texture of the shell. She said there was some evidence that the shell takes on a different texture if it has a disease. She asked, "Do you think your machine can measure it?" I bought some eggs. I brought them over to the lab. I measured them right here and they had the most interesting looking surface. I don't know if the project ever got off the ground. She might have bought a machine. I don't know.

How much of a part does reliable instrumentation play in helping overcome an ever-changing marketplace?

Here at Michigan Metrology, I have what's called a Bruker NPFLEX, which is a 3D optical profiler that allows measurement of large parts and small parts, both in very smooth situations like a super polished mirror all the way up to very rough parts, such as a brake rotor or something that's been shop blasted. That's the one instrument I use for surface metrology. Then together with Bruker, we've set up a lab that has a Bruker UMT, their Universal Mechanical Tester, which allows the measurement of friction and wear of components by rubbing them together in different configurations, be it a pin on disk, four-ball test, or whatever kind of configuration you want. Throughout the year, Bruker experts are here demonstrating the machine and doing projects for people, as well as developing applications for different people in the auto industry.

Bruker's NPFLEX

I've had the luxury of being involved in this from the very beginning, in the very first 3D optical profiler, which was called the RST. It's gone through many generations, each time gaining more and more capability. The NPFLEX pretty much has everything anyone could ever ask for. Given its ability and flexibility, you're able to measure larger parts. We're able to move parts in and out pretty easily. We can easily change the turret and use long working distance lenses, so for imaging large parts, the whole architecture is just a lot easier to use. The machine is also very fast compared to the earlier machines, both in its scanning space and the basic data handling. I find it much more efficient using the machine. I can get projects completed a lot quicker with the NPFLEX than other types of machine. I've had my NPFLEX for over three years now, and I've been using other optical profilers for about 30 years.

Learn about the instruments in Don’s Lab

About Don Cohen

Dr. Cohen has an undergraduate degree in Physics from the University of Michigan – Dearborn and graduate degrees in Physics and Optical Sciences from the University of Arizona. Early in his career, Dr. Cohen worked with IBM on optical disk drive development. He later joined WYKO Corporation as Product Manger and finally Vice President, developing 3D surface texture metrology instrumentation. In 1994, Dr. Cohen established Michigan Metrology,LLC to help engineers and scientist solve problems related to “leaks, squeaks, friction, wear, appearance, adhesion and other issues”, using 3D Surface MicroTexture Measurement and Analysis.