What causes transmissions to fail?
 

Sorry if this has been discussed before. I was not able to find anything talking about it.

I'm mostly wondering if transmission break when your shifting gears, or just from torque itself? If it goes from seeing 0 ft lbs to 250 when you are shifting. I'm assuming a transmission will see 0 torque for a second when shifting gears. Then would putting it into gear and pushing with the 250 ft lbs kill the transmission, or is it simply the ammount of torque applied? If you were to do a pull in 3rd gear and never shifted to 4th; would that still kill the transmission?

I'm just trying to get a better understanding of transmissions.
Thank you.

Don't know much about this topic but you may want to state which transmission you're interested in. Pretty sure I've heard the 6spd will take more than the 5spd.

Just in general. I don't even own a Miata, and don't plan to for a while. I currently own a saab 92x aero with a built ej257.

Op, clear you inbox. I'm trying to PM you.

done

The answer is both.
Shock loading, as well as sheer torque, as well as case flexing/shafts separating, etc etc etc.
There is no one answer to rule them all.

Do you think they are equally a factor in the cause? Let's say for the Miata's 5 speed.

I'm no pro, I think mathewdesigns is the token transmission guy around here, but here goes:

In order of destructiveness, I think shock loading (hard shifting) would require less torque to break the trans than sheer torque in a steady state pull.

My personal experience with a miata 5 sp supports this too: I babied it at 280tq for almost a year, mostly just smooth 1-gear pulls and even when I shifted fast it was always very smoothly and softly, then sold the car and the other guy broke it in a couple months.

In addition to the type of breakage mentioned above, I have been told by a transmission guru that the vast majority of "bad syncros", or grinding on certain gear changes, is not from the syncros but from bent shift forks. If the trans is warmed up you can shift very fast as long as youre guiding it in smoothly, not just slamming it into the general direction of the next gear with force behind it.

I often shift fast, but I never shift hard.

I would agree with 18psi. If you run enough power it is completely possible to break the tranny with just sheer sustained torque. Bbundy has broken several transmissions during 4th gear pulls on a straight away. The actual cause for failure is usually metal fatigue that results in a gear tooth sheering and fucking everything else up. That or a shift fork breaks.

Very useful. Would you say that the estimated 250 for the 5 speed braking was the results of hard shifting from owners or just that these transmitions can't hold the torque in general.

Would you say a turbo bp break the transmition faster than a vortec bp?

Power through a gearbox also affects wear and lube temperature. There was a thread on XMSN and Diff coolers, but I don't know who runs them or what they run.

The failures on this forum, however do seem to be more torque than power related.

Torque.

High rpm is a factor too but all mine basically have broken when the engine is very near its peak torque RPM in the middle of a pull. The RPM’s of the assembly is a factor in that it is what supply’s the cyclic loading that causes fatigue of the gear teeth. I could drive all around the track in one gear and still break a transmission. And I have never bent a shift fork or even wore out a syncro.

I'm going to say that the estimated 250 is assuming hard driving. Someone who is ham fisted can break it with less torque. Someone who is good at shifting fast but not hard can make it handle more torque. Then of course there's variance between transmissions, age of trans, etc.

I expect there's more to it as well, such as the spoolup time. Martin broke the 5-speed in Laz after 2 track days at 235 lb ft, but it was a super quick-spooling motor, and he drove it hard. Plus track = higher heat.

+1 to pretty much everything already posted.

The only thing I would say otherwise is that a misadjusted clutch is typically the main culprit in wearing out synchros (unless they have a lot of miles on them), not necessarily quick or hard shifting. If the clutch is dragging with the pedal fully depressed then the trans can't be totally disconnected from the engine, and as you are trying to slow the gear on a shaft you have to fight against the engine as well...the engine is going to win and cause a lot of premature wear on the synchro. Once the synchro is worn you start having problems with shifting smoothly at any speed, which precipitates worn/broken forks because of frustration and impatience :)

Is there a way to test if it is dragging?

Wheels up in the air, clutch in, put it in gear?

Two tests:

General shifting: Car on a flat, level street or parking lot. With the car running, shift into 1st and give it some gas (1/2 throttle is fine) with the clutch still fully depressed. If the car pulls forward the clutch is dragging.

At speed/WOT: On a highway get the car up to speed. Shift into neutral and leave the clutch depressed, go WOT, bouncing off the limiter, and at the same time try to row through the gears while holding the clutch pedal down. If it fights you in this test then it's still misadjusted but not nearly as badly as if you fail the parking lot test. You either need more adjustment or know that you can't shift cleanly at redline. The larger a clutch's diameter the worse this is, due to flex and vibration at the outer edge where it's farther away from the fixed hub.

Time to go test lol.

Load changes with rearend gearing as well. As soon as we dropped from 4.10s to 3.63s, the 5spd broke.

To add slightly to the miata 5spd discussion:

The 5 speed teeth shear off the ring. Most failures look like this, regardless of which gear failed. The metallurgy of the gears/teeth themselves IS a major issue. Simply, it's not designed to pass such a load through the gear set.

Shock load, total torque, rough shifting, and abuse, are all agitators, of an inherently weak design. Well, weak for a part designed for literally half of the WTQ numbers at witch failure becomes common place.

Another factor is clutch clamping force. A clutch which is slipping slightly will consume energy in that process. Additionally, an aggressive pressure plate and friction material will quicken the process of applying load to the gear set, thus contributing to "shock loading" forces.





I will agree with what's above, most "bad syncros" are issues with the linkage or the forks. Having truly damaged syncros in a modern manual, is fairly rare.

I hope this is the case. I'm putting in a 4.778.

I should note, that's 2nd gear in autocross. You'll still be loaded in high gear.

4.778 will be used with a Quaife wide 5speed and a 8600 rpm rev limit for atocross use of second gear. 3.909 will be used on tracks.

That should help you, Bob.

The one thing that hasn't been mentioned so far is the grip the car is running. A 300wtq miata on star specs is a very different beast than a 300wtq miata on 275 hoosiers on 11's. The latter case is much more likely imo to shatter a gear than the first. The engine is relentless, and the grip is relentless, and the transmission gives up in the middle.

What are the thoughts on gearbox temp vs life in high torque applications? Has anyone who's broken a transmission installed a temp gauge to see what the oil temps were like?

Transmission life went down big time when I switched from 225 Nittos on 9"to 245 Hoosiers on 10" for the track. I've only broken second gear once in the 6 speeds. It is usually 4th that goes. 5 speeds will break 2nd 3rd or the input gear that takes out everything but 4th at about equal probability.

I have also learned recently after taking a bunch of transmissions apart the NB 5 speed has wider gears throughout than the two versions in the NA particularly noticeable on the input gear. An NB 5 speed is most defiantly going to be a bit stronger than one from an NA. The 6 speeds have much thicker gears than the 5 speed as well but the 5 speed has 3 bearings per shaft instead of 2 which I think might be a design advantage to resist shaft flex. 6 speed was ~12 lbs heavier than the 5 speed Quaife with NB nose, bellhousing and NA tailstock I put together most of that weight I think is in the shafts and gears.

@Pat

The short answer is: Higher temps break down fluid which affects its ability to maintain surface film, and regardless of TQ numbers this will accelerate wear on the tooth. Higher TQ through the gearbox will increase that wear accordingly.

Bob destroys them in autocross, I don't think there's time for the gearbox to get very hot. I could be wrong.

Ok so based on that, would a good transmission cooler and keeping high quality fluid in it improve reliability in high torque applications? I'm considering installing a trans cooler on my 6 speed.

If you are tracking, I would think yes. There's definitely heat rejection advantages for track. For autocross, I can't recall any threads on this board with instrumented readings of trans temp, but given the relationship of other known cooling mods, autocross heat <<<< track heat.

Bob is breaking trannies because his car is overcoming the physical strength of the gears. The teeth are just not strong enough to hold the power.

A trans cooler will help as long as the engine is not going to be a monster like Bob's. Keeping the fluid at an optimal operating temp is always a good thing. Not only does the cooler keep the fluid cool, but keeps the entire trans colder, helping with any wear/breakage problems related to heat induced fatigue of the gearset and case.

I don't know for sure what my trans temps have been but I have never seen burnt trans fluid. always looks clean except for broken off gear teeth.

IIRC Bob breaks trannies at autocrosses, where tranny temps should be a non-issue. I used to think that a fluid cooler would help improve shaft alignment by keeping the bearings cooler, but I think breaking trannies at an autocross debunks that pretty soundly.

When there's enough power to overwhelm the gearset in short courses like an auto-x where no heat is generated then there's no cooler that will help. But in a lower TQ car (not enough TQ to shear teeth at will), I think a cooler would be beneficial if it is driven hard on track where heat generation will be a factor based on the points I mentioned above.

That being said, I don't personally have a track-dedicated Miata, let alone a small fleet of them, nor do we have any Miata customers at this time. If Andrew or Emilio or any of the other serious track guys are not seeing any benefit from a cooler then that's got to be considered as a hard data point.

Bob has also broken them in straight line under load on the track. So not even on hard starts. Last one I read about was 4th sheared off on the back straight of PIR or pacific raceway.

Yes. I think it's been established that Bob's engine will destroy gearboxes regardless of heat or shock load. He has insane grip and tons of torque.

Some have grenaded gearboxes on the track at considerably lower torque load, no? I would think that a trans cooler could really help a 6-speed live behind, say, 275 wtq instead of Bob's 300+.

Total conjecture, though, as Matthew said.

In current level of grip form a 6 speed has not lasted for more than two days at PIR. 4th gear goes on the back straight out of a corner taken in 4th gear so no shift is involved. I do hold the overall Miata lap record around the track by a good bit however and still holding on to the TT2 lap record. Nobody even got my Pacific raceway lap record even after repaving turn 8 last year helped most people drop about 2 seconds.

I also don't run full boost on the track I limit it to right around 320-325 hp ~ 285-295 ft-lbs though.

Have you broken a 6spd autocrossing?

Just one.
Broke second gear headed for the finish shoot. I run full boost autocrossing but if I datalog a run I never seem to see full load. Can't load the engine that much in second gear its either accelerating too quickly through the RPM range or spinning the tires. I mainly crank up the manual boost controller cause it seems to build boost quicker at lower rpm when I do and it feels more responsive to throttle inputs. I might be able to do better with electronic boost control but I have never gotten one good enough to be consistant and smooth and repeatable like the manual one. on the track I will see sustained periods of full load through several gears through powerband rev range. Autocrossing is a spastic looking thing by comparison.

Do you use an IWG or an EWG? I'm really impressed by the ease of boost control tuning using the Turbosmart EWG on my new FM2R (with 2863) vs the IWG on my old 2560.

--Ian