Dave,
Man, you have the best email structures I've seen in awhile. I wish I could organize my thoughts that well. It takes 5 revisions to get me where I'm somewhat organized in my thought-to-writing process. Well done sir.
Now, back to the "problem" at hand.
I was there when the JSM crew tore apart Jeremy's transfer case and looked inside the Viscous Coupler. If memory serves me, there is a picture floating around I took of Jeremys truck on cinder blocks in Jeffs driveway....Anyway, I would imagine I'm one of only a few who have actually seen the internal workings of the unit, considering it is non-serviceable otherwise. I feel I have a good grasp on it's workings.
I spoke with my professor yesterday. He has lead my classes on Mechanical Dynamics, Thermodynamics, Machine Elements, and Design Specifications (all of which I've passed with an "A" grade in the last two years
). He is also a licensed Professional Engineer in the state of Indiana. Here's what he had to say.
- You're right on track with the viscous fluid being the first fluid to heat up. The ultimate goal would be to keep that fluid cool, however since the viscosity of that fluid is so high that isn't practical. The next best thing, exactly what GM did, is to put that entire unit in an oil bath. He mentioned a water bath would be even better due to the higher heat absorption rate of water, but that was impractical.
- Since the entire unit is completely submersed in fluid, or rather has fluid circulated over it by way of the internal chain "slinging" the fluid around, the heat generated by the shearing of the viscous fluid will be transferred to the oil bath fluid. This transfer rate will be high at first, and slow as the temperature difference between the fluids decreases.
- Ultimately, the heat generated by the shearing of the fluid has to go somewhere. There are several transfers happening, and to quantify them all would take exhaustive calculations. Conduction out of the ends of the coupler to the shafts, as well as convection within the fluid, then to the VC housing, then to the t-case fluid, then to the t-case housing, then finally to the surrounding air.
- His advice to determine the maximum allowable differential between front and rear was this; make a best guess, then go test. Test a LOT. Then apply a safety factor to avoid warranty issues, and write the spec.This sounds essentially what Audi has done with my wife's '06 A4. The manual specifies a maximum 5/32" tread wear difference between any two tires. Audi, calculated, tested, then spec'd the allowed difference. My professor's guess was that GM applied a hefty safety factor to the design, never wrote a spec, and essentially "crossed their fingers, and rolled the dice". This is shown from the number of users that "Haven't had a problem in years", as well as those that have experienced "immediate failure".
- He commented after I explained the debate that generated the question was this; "It sounds like there are already many people testing the coupler for GM. Some have found the limit, others have chosen to stay within the obvious low risk zone. Only the owner can determine if the problems are (as we called it) "negligible". "
- He then followed all of this up with the question on the cost of replacement. I couldn't put a good number on it off-hand in our hallway discussion, so we assumed a $750 complete replacement of the transfer case. His thoughts were that if a person is really that interested in running an offset tire size for either aesthetics or performance (although we both agreed that the offset tires sizes were a hindrance to performance), that the price of a semi-regular transfer case replacement should simply be figured into the cost. There was agreement that perhaps a 40-60k miles expected lifespan could be expected of the t-case should one daily drive it with a somewhat short commute, and that $750/40k miles might be worth it to some, and not to others. An interesting way of looking at it if nothing else.
- We were both curious if you had the chance in the future, to measure the torque it would take to create a 10.26 RPM differential within the coupler. Essentially you would need to apply whatever torque necessary to move the input 90 degrees of rotation in about 1.5 seconds or so. That would be the empirical data we would need to do the calculations at 60 mph with a 1% difference in tire size.
So there we have it. Problem of the Sy/Ty world #2,794,103 solved. Ultimately I think we've put a decent amount of energy into this problem, only to land about where we started. It's possible, but at not-completely-known expense. Negligible to some, unacceptable to others, those are subjective words that will apply to each persons situation differently. Hopefully the reading has been at least somewhat interesting.
Secondarily; I'm going to post at least some of the conversation on .net. We've gone through a good deal of math, as well as enlisted the help of a well qualified consultant, and I don't want to essentially "horde" the data to ourselves. I think it's worthwhile to put it out there, even if only a select few "get it". I'll weed out your name and any identifying data of yours if you like, but I think the information should still be shared. Let me know what you think, and I'll act accordingly.
Have a good Tuesday sir!
Mike
