Aero momentum website implies inflight testing is more important. Doesn't list ground test hours, so I haven't found any industry standard yet:
"From this point we have spent over a year testing both in the lab and more importantly in the real world."
I keep preaching on this. To have a powertrain program be successful several things are essential:
- Well schemed out and well analyzed designs so that fatal flaws are eliminated, high value failure modes are protected against and eliminated where possible, making the design likely to be durable;
- Thorough bench and dyno test programs designed to expose weak spots, high value failures, design defects, and flaws in manufacture, and then demonstrate life targets are met. You tear down your stuff after it has run and see if anything is showing wear and tear you did not expect. You look for damage, pitting, wear materials, incipient failures, etc. Then you fix the things that you find wrong;
- Then operational testing, because no bench test ever covers the exact conditions of actually using the product. And you repeat the skepticism about how it is doing, looking for wear materials in the oil, tearing them down to see how internals are doing, and so on.
Mark at AeroMomentum has a valid point. He has designed his stuff to be sturdy in the real world. He has run his powerplant systems on stands for bunches of hours. He has established his operating parameters and envelope. But how does one simulate spin entries or abrupt rolling and pitching on a stand? How does one find out how long it actually lives with someone flogging it or babying it with modest power and short runs? How does one one find out how it does in Alaska and the Caribbean and in deserts and jungles? Go fly it and see how it does. If you have done the job up until now, the customers using your Beta product have little in the way of problems, and you make a big effort to solidly fix what remains. Ultimately though, the customer runs the product that comes from the shop way more and in more varied ways than some folks will even believe. That is where one finds out if the product is really good or just a pretender. No way to avoid this, but if a good job was done designing, engineering solutions to any remaining issues, and knocking them out, it is a pretty ethical way to do business. Not perfect, but if done well, it can be pretty darned good.
In fairness to Mark, he already has a lot of good customer time on working airboats running hard. That should count for a lot compared to whatever he has on a bench, and also validates his design/analysis/test processes. The thing that I keep emphasizing is that design that is well based is hardly a guess - one KNOWS where your analysis puts you on all the critical parameters. Then one tests that product and see that it proves reliable and durable or adjusts the product until it does. Once one knows that, the parameter set on future products must also be as good or better than the already proven one. This is engineering the next product.
Now if instead one starts with a set of guesses, and do not know where things like stresses, temperatures, etcetera are, and just go test, then bump materials or sizes or coolers until it passes your base test, one still does not know how well it will last or if the guys doing their thing in the field will just shred it. And one has no basis for making the next product decent from the start.
The issue of "testing on the customer" comes about when one builds too few test pieces (amazing how often sample sizes of one are accepted) with guessed at parameters, run it a couple hours on a stand, inspect and fix the biggest issue you find, run and keep fixing things on that one powerplant for a few more hours, and then decide to market and ship a bunch to unknowing customers so it can launch on time and make some money now. Is the customer protected a reasonable amount? Done this way is fundamentally unethical. Will the outfit grow and prosper? I personally hope that this producer is punished by the customer base.
Billski
