Monday, March 8

Composite Inspection

THIS is the question that prompted this whole series. CITIZEN RIDER was frustrated at a lack of solid information he can use. I suspect he’ll remain somewhat frustrated reading this, but he’ll be smarter and frustrated, which I hope will represent an improvement.

Basically, there are five major ways in which carbon composites are widely inspected today. Only three of them are likely to help a bike mechanic or owner wondering if that carbon rim is still OK after the curb hit.

These are:
Visual Inspection, Ultrasonic Inspection, Radiographic Inspection, Tap Test, and Load Test.

Visual Inspection
The first inspection technique has the major advantage that no special equipment is needed. Purchase price of the equipment is zero. Unfortunately, in many cases, you get what you pay for. Visual inspection is widely used as a part of the inspection process by the OEM manufacturers and it helps them immensely. I’ve often relied on photomicrographs to determine laminate quality and ply orientation of composites, and excessive porosity or resin pooling in the laminate is often evident to even a casual inspection. A bike mechanic or owner, on the other hand, is trying to determine if an impact is something to worry about. In the case of solid laminates, such as brake handles and such, a visual inspection may be very helpful. Carefully look for broken or loose fibers. Use a good magnifying glass in the area of the impact, and even more, on the back side of where the impact occurred. Damage is often more severe on the back side of a laminate. Still, even if you see damage, all it will really tell you is that you have cause to worry. In the case of hollow laminates, such as frame tubes, you are even in worse shape. The back side of the laminate will not be inspectible. If you are relying on visual inspection, the OEM advice to throw it away if you see something disturbing is about all that can be done. Still, looking is free and it can do no harm. Look hard, look REAL hard. Get a good magnifying glass. Inspect both sides of the composite.

Ultrasonic Inspection
Ultrasonic inspection is the inspection method of choice to detect and characterize delaminations and impact damage in composites. What’s more, it is the workhorse method for OEM manufacturers, at least in the aircraft industry. Equipment can range widely in price. On eBay, for example, an ultrasonic hand inspection unit can be obtained for somewhere in the $4000 range. Ultrasonic thickness units can be had on occasion for more like $200. I do not know if such could be used An aircraft manufacturer would use a more sophisticated automated unit that will be more sensitive. Such would be beyond even a very upscale bike shop, though some manufacturers might use such units if they were volume producers. An ultrasonic inspection works by detecting how far the ultrasound travels into the composite before being reflected by an air gap or other discontinuities that reflect the ultrasound. Essentially, it can “see” if the plies are separated internally or if there’s a busted up inside surface on that tube. You’d actually have a GOOD chance of finding it if you were a trained ultrasonic technician and had comparison standards to use. Some of those guys are wonders – equivalent to the Navy sonar operators that can tell whales apart from each other from miles away. Maybe there’s a business opportunity here for a savvy guy that wants to sell inspection services to bike shops. Still, even with the best ultrasonic operator, my engineers and I constantly puzzle over what the strength effects really ARE of that “indication” that was detected. Still, if nothing else, you’d have more than just a hunch before deciding to throw that $3000 frame in the trash. You could tell the customer that your ultrasonic inspection service got an indication of a half inch defect on the chain stay and that isn’t covered by warrantee. He could even frame the inspection report if he were so inclined.

There are a lot of places that can do ultrasonic inspection. A FEW are listed below. I've used none of them and cannot recommend any of them, either as being wonderful or useless.

PTL Testing
NDT Specialists
191 others at ThomasNet
One problem ANY of these places will have trying to inspect your bike is that none of them will have usable STANDARDS. A STANDARD is a reference part that shows what a GOOD part and a BAD part look like. All they will be able to tell you is there may be something odd inside that fork.

Radiographic Inspection
Radiographic inspections are widely used by OEM manufacturers. They are helpful in determining where internal porosity is in a laminate and if the fibers have bunched up during a cure. Besides being impractical for an end user, nothing an end user is likely to do will change anything a radiograph will find. The one time you might want to have a bike component radiographed is to establish evidence in a product liability lawsuit. The owner did NOT cause the composite to be porous or the fibers to shift to one side of a tube that failed. The bike was built that way. I presume most that are reading this post are not lawyers looking to sue Trek. Still, Radiograph is one way you can build a case that it wasn't abuse, but OEM failure to take due care.

Tap Test
Tapping is something I sneer at whenever someone tries to tell me it’s a valid inspection technique. Tapping, whether with a quarter or an official tap hammer has been repeatedly found to be somewhere between unreliable and equivalent to consulting a magic eight ball. The theory is that you carefully tap the laminate and listen for “dead” sounds. If inspecting a fork, you have the advantage that you can try tapping one fork leg, and then tap the other to see how they sound the same or different. If it’s starting to sound like dowsing, you’re not real far off. On the other hand, if I’m in the field and there’s a suspect carbon laminate, I’ll get my calibrated quarter out and carefully tap away. It’s SOMETHING you can do. As in the case of visual inspection, it won’t harm anything unless you rely on it. I prefer to use a new, shiny quarter. American ones have more “heft” than their Canadian counterparts. If nothing else, tapping the laminate will make you feel a little better.

Load Test
Load tests are, in my view, the superior method of establishing that a composite part remains safe after experiencing an impact or other post-manufacture damage. Ski binding settings are validated by a simple load test. You WANT a ski binding to release AT the intended load, neither coming loose during normal skiing, nor failing to release when needed. What’s more, load tests do not require specialized equipment or training for the operators. What’s best of all, if you load a damaged part to the proper test load and it splinters into a million pieces, there is no doubt about the proper course of action. You can tell that guy his $3000 bike was unsafe and the proof is in the bag of splinters you give back to him. What’s more, you can explain to him what the test is about and he can even watch. Because composites are not fatigue sensitive, if it passes, as long as it is a good load, you know the part is good to go. Myself, if I was running a load test, I’d cycle it a half dozen times, just to be safe.

If we were talking about an aircraft, conducting a load test on a wing would be a difficult undertaking. It is much simpler for a bike, where the loads are simpler and more “human sized.” In an ideal world, the bike and component manufacturers would provide directions on how a load test could be carried out and what the recommended test load should be. Even better, this would be provided by an industry group, since a safe fork test load on a road bike is not likely to be very different for, say Trek versus Schwinn. Ditto for that zinged chain stay. What’s more, conveniently enough, since there are industry and government standards, these could be made readily available for someone trying to decide to do with a possibly damaged component. Even in the complete absence of useful data from the OEMs, an owner need not be completely helpless. Even without me outlining it, I imagine most of you could devise a reasonable way to test a suspect crank arm. In the case of a front fork, it can’t do much more than throw you over the handle bars.

For reference, THREE locations I found that have at least superficial credibility are all by Sheldon Brown and Damon Rinard, namely here, here, and here. More on testing later...

I quizzed my engineers about all those inspection methods that CITIZEN RIDER identified. Mostly, I got quizzical looks back. The consensus was, however, that squid magnetometers sound really cool, even if they do nothing useful whatsoever. In the final analysis, however, the old truth still holds. “A test is worth a thousand expert opinions.”


Ham said...

As we don't have quarters, I was wondering if using a fifth (we don't yet use that term for 20p but maybe we should) will be 20% less effective?

Steve A said...

Remembering that I will sneer at anyone who takes tapping seriously in the first place, I'd favor a 10p coin or, even better, a 1 pound coin. If you are going to tap, you might as well get a solid coin. They give a more solid and consistent "thunk" when solid material is tapped. I'd stay away from the tiny coins and the ones that aren't round. Still, either way the numbers say we are fooling ourselves with coin tap no matter the denomination. At least the quarter in the photo was a TEXAS quarter and might impart some special power when used IN Texas.

So, Ham, do you believe in Nessie, too?

cafiend said...

I'm going to print this out to have around the shop.

Steve A said...

Make a note that YOUR worthless Quarter needs to say "Live Free or Die" on the back. None of that Lone Star stuff up there. Myself. I'm going skiing...

Post a Comment

No Need for Non-Robot proof here!