DIY Electric Car Forums banner

1 - 20 of 24 Posts

·
Registered
Joined
·
40 Posts
Discussion Starter #1
The reason for this post is that I'm storyboarding my project, rehearsing the steps before I undertake each one, and my thoughts on motor-trans alignment are completely in "How hard could it be?" territory. I am fully aware that thinking this way inevitably leads to ruin so I need to be set straight.

I know the importance of proper alignment, but what I don't know is why failing to achieve it seems to be such a chronic problem.

It seems to me that it's simple. Good measurements, some high school maths, and careful fabrication. Measure the physical dimensions of the two parts to be joined, calculate the relative positions of the mounting points and faces of the two parts when the shafts are coaxial and separated by enough space to allow for the connecting means, then fabricate an adapter that fills the space between the mount point/faces. Easier said than done, to be sure, but it's not rocket surgery.

What am I missing? Is the problem that the requisite precision and accuracy are more difficult than I believe them to be? Is it that too many people don't bother with the effort required to reach that level of precision and accuracy? Or maybe they get in a hurry and don't put in the time for careful forethought and planning?

Or is it that it is as simple as I think and most people are doing everything right but there are a handful of common problems that pop up again and again anyway?

Harmon
 

·
Registered
Joined
·
4,788 Posts
What am I missing? Is the problem that the requisite precision and accuracy are more difficult than I believe them to be? Is it that too many people don't bother with the effort required to reach that level of precision and accuracy? Or maybe they get in a hurry and don't put in the time for careful forethought and planning?
All of those are probably major factors.

Specifically for the first one: some people appear to believe that by making bolt holes in the adapter plate to match those in the engine and transmission, then putting bolts through them, all three parts (engine, adapter, transmission) will be aligned with sufficient accuracy. This belief seems to be most common among people who have never had an engine and transmission apart, so they haven't noticed the precision-fit dowels actually used to achieve alignment, or who don't understand the difference between a tension bolt and shear bolt.

By "careful fabrication", are you thinking of holding a hand-held drill really steady, or are you thinking of precisely operating a milling machine?
 

·
Registered
Joined
·
40 Posts
Discussion Starter #3
By "careful fabrication", are you thinking of holding a hand-held drill really steady, or are you thinking of precisely operating a milling machine?
Whatever it takes to hold to some reasonable tolerances. Take the time to build a jig to hold things if you need to, accept that hand drilling is fine sometimes but sometimes it can't be good enough, recognize when your drill press or bits are too wore out, banish "meh, close enough" from your vocabulary...basically, don't invalidate your good measurements with sloppy work.

Having access to a machine shop or is great if you've got it but I wouldn't think it is required to get good results for this particular job.

I'm certainly going to take advantage of having a machinist brother, but I'm worried I may be missing some important tidbits of info. It doesn't matter how good my AutoCAD prints are or how my brother takes pride in hitting his numbers dead nuts every time - if we're operating with bad info from the start it all goes out the window.

Harmon
 

·
Registered
Joined
·
727 Posts
Whatever it takes to hold to some reasonable tolerances. Take the time to build a jig to hold things if you need to, accept that hand drilling is fine sometimes but sometimes it can't be good enough, recognize when your drill press or bits are too wore out, banish "meh, close enough" from your vocabulary...basically, don't invalidate your good measurements with sloppy work.

Having access to a machine shop or is great if you've got it but I wouldn't think it is required to get good results for this particular job.

I'm certainly going to take advantage of having a machinist brother, but I'm worried I may be missing some important tidbits of info. It doesn't matter how good my AutoCAD prints are or how my brother takes pride in hitting his numbers dead nuts every time - if we're operating with bad info from the start it all goes out the window.

Harmon
Talk to people at a transmission repair shop. They'll probably be able to tell you about the importance of good alignment, in this situation. Think about the OEMs using the alignment dowels, in the first place. They've spent a lot of time and money acquiring and maintaining the equipment to accurately machine the holes for the dowels in their products. If the less accurately machined(and cheaper to make) bolt holes were all that were needed for proper alignment, why would they go to all of this trouble and expense?

Talk to your brother, the machinist. Tell him the true center lines of the transmission input shaft and the motor output shaft need to be parallel (the easy part), and off-set no more than ~0.005"(~0.13mm) (the hard part). See what he comes up with.

As for as "bad info", that's the basic problem of garbage in, garbage out.
 

·
Administrator
Joined
·
6,117 Posts
Hi
My tuppence worth is a lot of (most) of the "alignment issues" are actually issues with trying to use a horrible key type shaft in an application that it is entirely unsuited to
 

·
Registered
Joined
·
40 Posts
Discussion Starter #6
Hi
My tuppence worth is a lot of (most) of the "alignment issues" are actually issues with trying to use a horrible key type shaft in an application that it is entirely unsuited to
Interesting. I wouldn't have thought the fact the shaft was keyed as being the problem. I would have pointed to poor fitment of part being attached to the keyed shaft as the real problem rather than the nature of the attachment.

What is the better alternative, given that a DIY'er won't have the skills or tools for spline cutting?

Harmon
 

·
Administrator
Joined
·
6,117 Posts
Hi
I have a splined end on my motor

You should NEVER drive through a key - instead you want to drive through friction between the shaft and the adapter
That means that you need to CLAMP the adapter into the shaft

If I had a keyway I would

Tapp a thread in the end of the shaft
Get an adapter that seats down on a shoulder
Use a bolt to clamp the adapter down onto the shoulder

Or use the type of adapter that closed in squeezing and clamping the shaft

The aim would be to get enough squeeze/clamp so that the key was unessesary
I would probably keep the key as a backup

The key is not only not a good way of driving but it can make the adapter eccentric to the shaft

If you have a keyed shaft then you may not be able to get the adapter totally concentric
If so then you NEED to reduce the whirl loads - the sort of loads that you get with a flywheel
So with a keyed shaft I would abandon the idea of a clutch - or at least machine the flywheel down to almost nothing
 

·
Registered
Joined
·
217 Posts

·
Registered
Joined
·
727 Posts
Very Interesting , and to go slightly off at a tangent. Ive been doing work on Iveco 4x4 transfer cases lately , The center sandwich plate has spherical mount bearings - Meybe to account for misalignment issues.
If you're writing about spherical outer race bearings, they work great for a shaft with two bearings. But, a typical front engine, rear wheel drive vehicle transmission may have up to five bearings on the concentric input/output shaft combo, if the pilot bearing and tail-shaft bearing are counted. There's no flex plate or U-joint on this combo shaft to make up for misalignment. To work properly, the bearings need to be in carefully machined, located, and rigid mounts. Hence the need for the carefully located and machined dowels holes and dowels in the bell housing to adapter plate joint(or the ICE in its original form)
 

·
Administrator
Joined
·
6,117 Posts
When looking at the measures that the motor manufacturers take to achieve alignment remember that they are in a different game

They have a requirement that the pieces go together every time as fast as possible and that after being bashed together the alignment is good enough so that 99.99% of the vehicles last over 200,000 miles

We can individually fit our parts - take a bit of time faffing about

And we are OK with a failure rate that is thousands of times higher

So don't look at the OEM requirements and think that we need to meet or exceed them
 

·
Registered
Joined
·
1,466 Posts
Anytime you install a new bell hosing the manual calls for it to be dialed in then pined . bought a bell housing from Advance Adaptors to use a different transmission . It was very clear that this needed to be done.
I worked for a equipment mechanic who replaced a a damaged bell housing
on a 18 wheeler. He did not read (maybe at all) the manual and did not dial it in. He could not figure out why it was going through clutches .We argued over not dialing it in. I was the bad guy.
My best friend , a engineer had no idea about this procedure .
My machinist teacher talked endlessly about the need for concentricaty
but never showed us how to achieve it.
Same with welding procedures . compliance is rare.
I explained this on the forum before
Now you have to center the bell housing to the motor in your case.
So take a large bearing race clamp it to the face of the bell housing ,adjust to 0 run out. . Place the bell housing in a lathe and adjust so the clamped bearing race reads 0 . Now your lathe and bell (or other part) is
ready to get the machining to fit the motor concentrically on this surface.
 

·
Registered
Joined
·
1,466 Posts
I left out the parallel part which is running the dial on the face ,then doing that again on the lathe.
On the lathe the dial goes on tool post.
 

·
Registered
Joined
·
40 Posts
Discussion Starter #15
Anyone want to proffer an opinion on using giubos for coupling motor to trans? They're used in some pretty high HP applications so it seems like a perfect fit in this situation to absorb shock, isolate vibration, and accept any minor inaccuracies in alignment.

Harmon
 

·
Administrator
Joined
·
6,117 Posts
Hi
That would work as long as the gearbox input shaft is already supported in two bearings

If the gearbox input shaft is designed to be supported at one end by the motor/engine then you wil wreck it pretty fast
 

·
Registered
Joined
·
40 Posts
Discussion Starter #17
Hi
That would work as long as the gearbox input shaft is already supported in two bearings

If the gearbox input shaft is designed to be supported at one end by the motor/engine then you wil wreck it pretty fast
What would be the mechanism leading to the demise? I understand that the VW Type I transmission doesn't have a forward bearing for input shaft support and depended on the gland nut in the end of the crank. However, the situation is somewhat different in this EV instance.

In the ICE application the input shaft and motor crank moved independently of one another, meaning that the input shaft effectively floated in space without support. The clutch disk and flywheel could become non-concentric and the motor could whip the off-center input shaft around, destroying everything. The gland nut ensured that the axis of the trans input shaft and axis of the crank were centered on one another. The pilot bearing in the gland nut never saw the engine power, just deflection forces from imbalance in the system.

In the case of my EV application the guibo would keep the motor shaft and input shaft turning at the same rate and centered on one another, effectively locked to one another in terms of rotational speed and axial alignment. It should provide all the 'support' needed by the input shaft since the pilot bearing in the ICE crank wasn't about load bearing as much as it was about location. The guibo would confine the input shaft deflection to a few hundredths, easily within tolerance of the input shaft seal.

No?

Harmon
 

·
Registered
Joined
·
4,788 Posts
Anyone want to proffer an opinion on using giubos for coupling motor to trans? They're used in some pretty high HP applications so it seems like a perfect fit in this situation to absorb shock, isolate vibration, and accept any minor inaccuracies in alignment.
I had to look up "giubo"... wow, I thought that these things were long gone from automobiles. They were popular as cheap flex joints for axle shafts of rear-drive cars with independent rear suspension, in the 1950's. They couldn't handle much power, and didn't allow much suspension movement (angle change). On the other hand, in a propeller shaft the angular limit and torque limit would both be less problematic.

A quick search showed that BMW used these at least into this century, and may even still use them. BMW does a lot of strange stuff, including using semi-trailing arm IRS - state-of-the-art for the 1970's - up to 1994 on premium-priced "ultimate driving machines".:rolleyes:

It might be a German or British thing... apparently even the current Ford Transit van has one.

That would work as long as the gearbox input shaft is already supported in two bearings

If the gearbox input shaft is designed to be supported at one end by the motor/engine then you wil wreck it pretty fast
I'm not convinced that either of these scenarios are a happy place for a giubo coupling.
  • With a transmission input shaft on two bearings, the giubo would need to accommodate non-concentric shafts. That's not what the giubo is for - it accommodates intersecting non-collinear shafts (slight angle at the coupling).
  • With the transmission input shaft supported at one end by the giubo coupling, the input shaft would be located only by the rubber. That may be okay for the input shaft seal, but is likely not adequate for the set of gears on that shaft... although that's only an educated guess.
 

·
Registered
Joined
·
4,788 Posts
What would be the mechanism leading to the demise? I understand that the VW Type I transmission doesn't have a forward bearing for input shaft support and depended on the gland nut in the end of the crank.
...
In the case of my EV application the guibo would keep the motor shaft and input shaft turning at the same rate and centered on one another, effectively locked to one another in terms of rotational speed and axial alignment. It should provide all the 'support' needed by the input shaft since the pilot bearing in the ICE crank wasn't about load bearing as much as it was about location. The guibo would confine the input shaft deflection to a few hundredths, easily within tolerance of the input shaft seal.

No?
The load borne by the pilot bearing is all radial (very little axial, no torque), but I would guess "no" on the adequacy of support, because the radial location tolerance of the rubber giubo coupling is much looser than that of the pilot bearing.
 

·
Registered
Joined
·
217 Posts
Guibo, AKA Rotoflex or Donut Joint , Common use well into this century.

Used instead of universal joint in LandRover Discovery rear driveshafts .

And Holden Commodore Driveshafts , AKA Pontiac G8 . check it out.
 
1 - 20 of 24 Posts
Top