Srtaight Cut drops V Standard Helical ?
Turbo Mini Forums

Hmmm, the damage in mike1098's case may be more due to lack of oil than all the issues we are debating.

But, I don't want to be negative on any theories being put forward - I don't have any expertise on geartrain design - but I would like to improve on what I have with minimal cost/invasive work, even if it's not the "ultimate", just so long as it's an improvement. EDIT, ie, an improvement on helical gears so I don't have to use those boyracer noisy straight ones.....

Edited by Rod S on 10th Oct, 2008.

Schrödinger's cat - so which one am I ???

Guys, the needles don't have to 'fail' or have slop to create the thrust - remember the idler can 'slide' along the needles infinitly. Draw a FB diagram if you want to understand it better - you can simply apply a single axial force to simulate the total axial and another for the radial.

I feel the erratic-ness of the failures of the alloy faces is surely pointing more to lubrication than load? Otherwise every 998 A+ would be perfect at the idler thrust faces in the housing - and we know this is not neccesarily the case.

Dave,

Once upon a time I could draw an FBD but now I would have to go and get out my old Uni notes and books and re-teach myself.....

I tried doing it simplistically a couple of days ago with the helical side loads and fooked up bigtime.

Bottom line - in your opinion (as in working in that industry), is it worth us persuing improvements to the thrust arrangement alone (esp. if we want to use helical drops), or do we need to look at the radial arrangement as well ???

Not trying to put you on the spot but I would value your opinion.

Rod.

Schrödinger's cat - so which one am I ???

I am not keen on having the onus thrust onto me for the outright answer, but I'd start with a pressure fed, open chamber (one that fills, then overflows the excess volume) feed for the bearings, which runs through a decent filter.

I'd start there (easy, achievable, and a good starting point for future development) and start putting duty on the bearings. The cleanliness of the oil is something that you simply can't put a value on - it is impreitive to improve on the single filter used for engine and transmission, especially when the transmission is using oil way less than it should be...

I'd also look at a way of supplying oil before the mesh of the primary and idler - you can drip, but at high speeds the oil will never make it into the mesh.

An oil spray (jet) would hence be good. This wouldn't have to be filtered to the same quality as the bearing oil - but if pressure feeding the bearing cavity, you may as well use the same for the gear spray.

BTW - DON'T spray into the mesh, this is the worse place pf all. Spray no less than 45 degrees before the mesh - 90degrees before would be better.

Over time, filtering the oil will also clean ALL the trans (and engine) oil, which will help no end on all bearings.

Magnets in the sump are always good - I'd also put small ones in the bottom of the cavity you are overflowing to feed the idler bearings as well...

As sprox pointed out in the idler post - making sure oil can get 'through' the thrust washer is a good call. I opened my thrusts oil-groove up to try and make this happen, but in reality the difference with SC drops is so little anyway, it might not be noticed.


So start with the above, take pictures, measure clearances, of all the parts before and after a period of running, and see if there is anything amiss.


Oh - you might also want to consider accurately measuring the 'front' thrust washers for wear - to see if implementing Carl Austins patented method of pressure-feeding the thrust washers could be of use to reduce the friction at this point.

paulH i thought you might find these interesting,thy are the drop gears from an auto min box . bigger bearings and gears ,presumably to take the torque multiplication of the torque converter .
also see how theres a sort of ledge inside the gearbox casing ,this creates a bath of oil that can only get out through the bearing ,bit of a constant feed of oil .
regards robert

Attachments:

37.42kb

35.74kb

30.26kb

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

umm thanks robert now youv got me thinking very intresting.

Holy crap guys, I wish I had known this thread had been going on, the title was misleading... Rod S just pointed me to it. It might be dead now but I wanted to add my two cents.

As far as my failure, I believe Rod S is correct in that it was caused by lack of oil. There was play in the needle bearings, but not a catastrophic failure.

As far as the forces, the axial loads definitely cancel each other out. However, there is a moment force that remains, that would cause a "twist" of the idler gear.

I am fortunate in that Timken relocated one of their Technology Centers to Greenville, SC. This means I have a bunch of neighbors that work for them. I had one of these Timken guys over to look at my issue. His first reaction was to fit a needle thrust bearing. I talked him out of it because it sounded too elaborate. At that time I thought it would be much easier to find a steel thrust washer of the thickness that I required... 0.155" thick, to make for the Al that wore off.

On a side note, I did find that the MGAs use a steel thrust washer for the layshaft in the gearbox that is .155" thick, however the ID is 5/8" and the OD is 1.25". I could potentially mod the ID to get to the proper diameter but I was not sure about the OD. Going from 1.75" to 1.25" is a pretty big reduction in bearing surface. I have fears it may start digging into Al case.

SO... my plans... I have started running the numbers to run a needle thrust bearing. Most likely that will be the path I go down. I will send my numbers to my Timken buddies tomorrow to see if they think they have a bearing that will work for me.

If people are interested I will let you know what I settle on. Unfortunately, at the rate I am going it may be a while before I actually get some miles on my engine.

Alright, I dusted off the brain cells and came up with the following force calcs.

Basically for my engine, which is a 1098 that will have a small turbo...I am guessing I will have about 100 hp at 6,000 RPM (very much a guess). Most of the forces cancel out except for the axial component. For my assumptions that would equate to ~516lbs per side. The resulting moment put on the idler gear would be ~133 ft lbs. Depending on how tight the axial float is will determine how much load is transmitted to the thrust washer vs to the needle bearings. The needle thrust bearing will have to be able to handle ~4,700 RPM if I keep the red line at 6,000 RPM.

If you can follow my chicken scratch...

Edited by Mikes1098 on 8th Nov, 2008.

I'll watch with interest....

I first considered needle roller thrusts but Sprocket pointed out speed limitations and when I looked up a comparable size one in my SKF catalogue, it would have been beyond it's rating. So it will be interesting to hear what your Timken neighbours think.

One point on your calcs..... From the age of all your other bits, presumably your drop gears/idler are "A" type, not "A+".

The helix angle is greater and the shaft is slightly larger on "A+", won't affect the principle, just the detail.

Schrödinger's cat - so which one am I ???

The only other problem to consider with the roller thrust bearing is how you are going to apply the correct pre load.

If you do not pre load the bearing, the rollers will slide rather than roll.

Moving forward… going to give the needle thrust bearings a go!

Potential Issues:

SPEED: The Timken needle thrust bearings seem to higher ratings than the SKF ones. The bearing that was recommended to me is rated for 14,000 RPM (assuming this is the kinematic limit), not sure what the thermal limit is but looking at the SKF ratings, the reference speed seems to be ~50% of the max speed... so with that assumption these bearings should be good for ~7,000 RPM. Keep in mind, they will be installed on an A engine not an A+, so the sizes, loads, and speed are slightly different. The max speed my idler will see is 4,645 RPM (which will be a rare occasion). SO speed will not be a problem.

PRELOAD: Not knowing how stiff the flywheel housing is, it would be difficult to nail a preload. I plan on shooting for the tightest clearance possible without intentionally preloading the bearing, even though a little preload would not be a bad thing. This bearing really only needs ~3 lbs preload (not much at all). Too much preload would only create unnecessary drag. The other thing to keep in mind, unlike the primary and drive gear which loads hard to one side, the idler actually will apply load to both bearings at all times…thus achieving the necessary loading. Obviously, if this were a production car I would have done a FMEA one flywheel housing to determine the flex and thermal growth with temperature to really dial this in…for my one-off special case, I am calling the preload good.

LUBRICATION: If the thrust washers had enough oil, which mine didn’t at one point in time but we will pretend they get enough oil, the needle bearing should be fine. They would require less oil than the plain bearings with rely on an oil wedge.

OD REDUCTION: No concern…bearings are rating for WAY more load than I am go to be sending it.
-------------------------------------------------------------------------------------------
I can’t find anything that would say this won’t work, so I now have bearings in my possession and they are thin and sexy! Once I saw them in real life, I felt even better that this might work.




This is the hardened washer that goes on both sides of the bearing. You can see the damage to the idler gear’s thrust surface from the washer failure. I have cleaned up the surface on a lathe… everything you see in low… actually looks way worse in the pictures.


You can see the OD difference from the original washer. These things fit beautifully…




I would be interested to hear people’s thoughts on this.

Mike

Edited by Mikes1098 on 8th Nov, 2008.

Very interesting...

I started off on completely the wrong assumption that there was a very large axial load to be dealt with so I was looking at the largest possible diameter bearing that would fit, ie as large as, or maybe even larger, than the existing washer.

That is most likely why I was finding speed ratings to be borderline or inadequate.

My false assumption is now well and truely corrected but I never went back to look at bearing sizes again.

What you have there is a much smaller bearing than I was originally imagining.

The fact that Timken have better speed ratings than SKF is a little surprising but obviously an advantage.

Pre-load, I agree with you (now I understand the loading of the gear) they should both be carrying some load all of the time provided you get the initial "clearance" right.

Lubrication, might be worth some more thought along the lines of Turbodave's comments. The whole flow regime around there will be different with an "open" needle thrust rather than the enclosed washer, it might be the cylindrical needle rollers suffer as a result.

OD, no concern from the bearing ratings but what about the practicalities....

The main thing that stands out from the photos is just how small a diameter it is compared to the original washers so, when your hardened washer (that makes the outer part of the bearing) goes up against the aluminium it is sitting on a much smaller area than the previous thrust washer did... What if it rotates relative to the aluminium ??? If it did, the wear rate would be much higher than the previous washer.

Have you considered anyway of locking it so it can't turn ???

And final practicality - how will you set the clearance ??? Additional shims behind one of the washers or just very accurate machining of the aluminium ???

None of that is meant to be negative, please don't take it that way, I'm just thinking of the practicalities of making it work.

Final question(s), what size bearing is it, is it imperial or metric, was it an exact fit on the shaft diameter... I then need to find out if there is an appropriate size for the A+ shaft.

Schrödinger's cat - so which one am I ???

mike , you excite me !

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

we will just draw a discreet veil over that one Robert

Edited by PaulH on 7th Nov, 2008.

These are small thrusts for sure - especially when you consider this is a pre A+ idler.
I'd consider counterboring the transfer housing slightly and adding a small belville spring to keep a pre-load if that really is essential.

On 7th Nov, 2008 PaulH said:
we will just draw a discreet veil over that one Robert

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

On 7th Nov, 2008 robert said:

On 7th Nov, 2008 PaulH said:
we will just draw a discreet veil over that one Robert

um.... yes . probably best :$ :)

Edited by Ben. on 7th Nov, 2008.

Build Thread Click Here

From the Crazy Idea department...the original thrust washers were up against the side faces of the gear. The shaft stubs sticking out of the gear have flat faces at their ends which could similarly take up some of the thrust load.

Edited by Mikes1098 on 8th Nov, 2008.

Good stuff....

I've downloaded Timken's technical catalogue - to my surprise it's freely available on the internet - and it's every bit as good as the paper SKF one I use, probably better because it lists all the imperial sizes too which my SKF one doesn't.

It's here for those interested, 3rd one down on the list
http://www.timken.com/EN-US/products/Pages...ctsCatalog.aspx

Regarding my concern over the outer washer rotating against the aluminium, it seems the answer, in an ideal world, is to pilot the outer washers in the housing, not on the shaft.

ie, the housings should have a small counterbore the size of the washer OD so the ID of the washer doesn't touch the shaft.

But that is in an ideal world...... the washers are available in several thicknesses, BUT, I doubt there is enough "room" to counterbore the transfer housing or gearbox case before hitting the cylindrical needle roller shaft bearing.....

I shall go and do some measurements though.

The real shame though is, they do a combined needle roller radial (shaft) and thrust bearing package that would overcome the potential problem entirely.... but only in metric sizes

From memory, I think the A+ shaft is 7/8" - here the thrust is available in two different ODs, both of which have adequate speed rating even though they are obviously bigger than the "A" size - but I need to be sure of the A+ shaft size.....

Anyone got one to hand to measure ??? all my A+ idlers are inside engines at the moment......

Schrödinger's cat - so which one am I ???

A small overlook that I found when assembling the thrust bearings... the bore for the radial needle bearings will leave the thrust washer unsupported for about half the length of the roller :(

I am thinking I have to go with thicker washers...how thick, I have no clue... that will require me to machine the case again :(

Maybe an opportunity to recess the outer washer(s) into the casings ???

Schrödinger's cat - so which one am I ???

Yeah, on the cover side...but the gearbox side is all built so it will go without the recess.

I WISH I HAD A MACHINE SHOP! or at least access to one....

I haven't checked the SKF and Timken catalogues yet but is there any reason that ball bearing thrusts have not been mentioned yet? I will read over the technical data on them soon but off hand they have a high speed rating. We used to use them as an equivelent to a clutch release bearing in radio controlled cars with engine speeds of ~50K RPM. They required very little lubrication in this application. Also, due to their low profile (OD - ID) maybe 2 could be used together with one outside the other. The only glaringly obvious disadvantage (to me anyway ) over needle rollers is the fact that they would be wider.

Opinions please

On 12th Nov, 2009 Paul S said:

I think Gary OS has taken over my role as the forum smart arse