Does my rear axle housing need to be straightened?

[Bill Robinson]

Does anybody know the limits on when an axle housing is bent enough to warrant attempting to straighten it? My axle shaft is off-center in the housing by .134”. This was a new, modern axle shaft that was installed in 2017. The shaft itself does not appear to be bent. The Hyatt bearing is good, within tolerance and no play, and the shaft itself has no end-play. What do you think?

[Norman Kling]

I don't know the maximum distance recommended before straightening, however, there is a bit of slack in the roller bearings. The main thing to be concerned about is that if there is not enough slack for the axle to find a way to run straight, it would bend a bit with every revolution and eventually would get metal fatigue and break. So I guess it depends on how much you drive the car how soon that would happen. Someone else might know the vital statistics.
Norm

[Bill Coyle]

Bill, the difference between your measurements is .134" but the shaft is only half that amount or .067" from center. Not that this answers your question of how much is too much.
Bill Coyle

[Charlie B in N.J.]

Hold on. Does that measurement really mean anything? Doesn’t the axle not sit centered in the housing without the outer bearing to keep it there? What am I missing?

[Russ T Fender]

Might not be the best advice but as long as the roller bearing will slip in without having to be aggressively forced in place I assume the housing doesn't need straightening. I usually try to fit the bearing four times with the axle turned to the 3,6,9 and 12 o'clock positions using the hole for the split pin as a reference. There might be a bit of a difference in the pressure needed at some points (which I assume means something is bent a bit) but as long as it isn't significant I go with it. I have seen plenty of housings where you would have to pound the roller bearing to get it to go into place and those I either put aside or straighten.

[Les Schubert]

I have straightened enough rear axle housings over the years! Fairly easy to do with just a oxyacetylene torch.
I made a set of testing fixtures and I will try to post some pictures

[Les Schubert]

Further to Bill’s posting, I think his is probably straight enough

[kmatt2]

I don't think that you can tell what is off that way. Did you measure the front and back also then rotate the axle 180 deg and check all four measurements again ? That would show if the axle is straight but it could be the housing or something amis at the differential carrier. I would say that if you can slide the good outside roller bearing over the axle into a good axle sleeve with out a lot of force, you should be OK.
The next time you have the rear end apart you can check the housings by their self. On a 1913 and later housing you can use a plumbob, clean cardboard, straight edge, a framing square, and a marker. Place the housing diff side down on the cardboard, place straight edge at pinion opening and draw a line along straight edge. Trace the outline of the outside of the housing center on the cardboard. With someone to help put the plumbob into the tube and center the stirring at the outer bearing opening. When the string is centered use the plumbob point to mark a hole in cardboard, remove housing. Now with a framing square draw a line at 90 deg from pinion line to plumbob mark. Now you can check how square and centered things are, any bend will be exsarated by the housing length. I haven't tried this on a pre 1913 housing they would require better measurements.

[Steve Jelf]

http://www.mtfca.com/discus/messages/17 ... 1297335103

[John Illinois]

I straightened a housing the way I have straightened V8 Ford torque tubes that were shortened. I mark the side of the tube that needs to be pulled up
near the bell. I run a mig bead parallel with the axle. I started about a one inch heavy bead.I use the axle as an indicator. If it needs more pull up run more weld.
Grind weld smooth and it is done. Not near as precise as the methods mentioned,but seems to work.

John

[RGould1910]

If it were mine, I'd straighten it. I don't believe there is any service limit though. With the system I use, I would heat shrink the axel close to the center housing.
On the other hand if the roller bearing slips into the end of the housing without binding I might leave it alone. I'm sure in the old days, folks simply substituted a worn roller bearing and let it go.

[Art M]

I have made extensive calculations on the axle bending, bearing misalignment, and bearing loading from the tire and gears. Your misalignment could be beneficial or detrimental depending on which way it is. Years ago car manufacturers offset to compensate for loading misalignment. That was 50 years ago.
I wouldn't worry about your issue. You could check the misalignment at the other end, but another good bare axle would be needed.
Art Mirtes

[Jerry VanOoteghem]

Is your drawing really correct in that it shows the axle sitting above the housing centerline? A bent housing would typically cause the axel to sit below the housing centerline.

Also, it appears that your inner seal is installed as you take your measurement. At some amount of deflection, the seal will begin to support the axle, giving an erroneous sense of just how much offset there really is. I usually check for any offset/deflection with the inner seal removed. Since there's always a bit of "droop" anyway, due to gravity, I lift the axel until I get some fair resistance, note how it's offset from the bore, then compare that to the offset when the axel is unsupported, (i.e. drooped). If the "up" & "down" offsets are the same, the housing is not bent. If they differ, then the difference suggests how much the housing is bent.

As someone else suggested, if you can insert the Hyatt bearing without using undue force to raise the axel enough to do so, then you're probably in good shape.

[Bill Robinson]

Jerry- the drawing shows the centerline of the axle tube is actually below the center of the shaft. I agree that the inside seal might be supporting (I hope!!) the shaft and maybe if I remove the seal it will let the axle drop approx 1/16" (could I be that lucky?).
In the morning, I'll go to my shop and pull the seal, and re-measure. My fingers will be crossed.

Meanwhile, I have some new numbers. This time I took the earlier suggestion that I should rotate the axle shaft and remeasure, to see if the shaft might be bent rather than the tube.

4 views of the axle shaft 90 degrees apart

[Bill Robinson]

OK. I removed the inner, modern cone shaped rubber seal. Luck did not go my way. Instead of the axle moving down toward the center as expected, the axle shaft lifted slightly to the top of the axle housing tube. The clearance at the top changed from .514 to .466.

[TRDxB2]

Hold on. Does that measurement really mean anything? Doesn’t the axle not sit centered in the housing without the outer bearing to keep it there? What am I missing?

I would also assume that to be true too that's why there are inner and outer bearings, right?. And the axle should not be attached to the load of the Model T. Also shouldn't the axle itself be measure to ensure its true - even though I doubt that would be the cause - it would reflect the method of the measurements.

[otrcman]

Bill, I'm probably the only guy who doesn't know, but what is the context of the picture ? Is the axle installed in a car or is it out of the car ?

If the axle is on jack stands (whether asle is in the car or not), where are the jack stands placed relative to the spring perches ?

Is just one end of the axle off the ground or both ?

Is there a wheel on the other end of the axle ?

[RGould1910]

The new measurements are too far off to ignore. I'd definitely find the cause and correct.

[Jerry VanOoteghem]

Bill,

While nothing is impossible, I would think it's fairly rare to have the axel tube bent downwards at its outer end, causing the shaft to sit above its centerline. I would think that housings get bent due to the spring bottoming out and smacking the center of the axel housing, making the whole assembly "swaybacked". What you're describing is the opposite, with the whole assembly being arched upward. Not personally having had a housing bent to any noticeable extent, perhaps my understanding of how the bend commonly occurs is flawed. Maybe others can offer their findings.

How much force is needed to push the axel back on center? Can it be done easily by hand?

As Richard suggests, it's probably worth looking into further...

[Bill Robinson]

Dick- the answers to your questions
Is the axle installed in a car or is it out of the car ? The axle is attached to the car per the leaf springs.

If the axle is on jack stands (whether asle is in the car or not), where are the jack stands placed relative to the spring perches ? Drivers side has the wheel attached and is on the floor. Passenger side, tire and wheel have been removed, jacked up, resting on a jack stand near the outside end of the axle tube.

Is just one end of the axle off the ground or both ? see above

Is there a wheel on the other end of the axle ? see above

[otrcman]

Bill, I have never actually measured an axle shaft or a housing with the axle bearing the weight of a car, so I'm willing to be educated. But the loads imposed by your car as it stands right now may be giving you erroneous readings. Most of the time when people measure axle shafts or housings for straightness there are no external loads being applied to the parts, especially unsymmetrical loads as your car stands now.

Typically the measurements that you are taking are done with the rear end removed from the car and resting symmetrically on two supports.

Your measurements may vary a lot if the axle unit is removed from the car and the other wheel removed.

But, I am willing to be educated.

[Bill Robinson]

Dick- initially, my problem was grease slinging out of the hub and onto my rt rear tire & brake disc, so I pulled the hub to check it out. That's when I noticed the axle shaft was not centered in the axle tube like it normally is. That's when I posted my thread on the forum. Now, let me say this- "it ain't my first rodeo", and I know that any rear axle repairs are normally done with the axle on the bench in a vertical position. This way, gravity will not play as much of a role as having the car on a jack stand. I also have disassembled many Model T axles and have never noticed the shaft being visibly off-center in the tube. That is why I asked for opinions on this particular problem.

[Jerry VanOoteghem]

Bill, I have never actually measured an axle shaft or a housing with the axle bearing the weight of a car, so I'm willing to be educated. But the loads imposed by your car as it stands right now may be giving you erroneous readings. Most of the time when people measure axle shafts or housings for straightness there are no external loads being applied to the parts, especially unsymmetrical loads as your car stands now.

Typically the measurements that you are taking are done with the rear end removed from the car and resting symmetrically on two supports.

Your measurements may vary a lot if the axle unit is removed from the car and the other wheel removed.

But, I am willing to be educated.

By Bill's description, the jack is located almost directly beneath the load exerted by the car and therefore should not be a factor here.

[Art M]

The axle housing should be straightened. I previously said don't worry about it but I was thinking .0067 inch instead if .067 in. off center. The hazards of not checking and rechecking. Art Mirtes

[Norman Kling]

It is possible that over the years, the car has been jacked up with a jack under the center of the axle. That could bend the tube so the outer end is lower than the axle shaft. Unless this is a rare car, you can very likely find another axle tube to replace it, or carefully bend it back. Another possibility would be the inner bearing is worn out which would cause a sag in the center carrier.

Norm

[TRDxB2]

Question: Not familiar with how to measure a bent axle housing as being done but... It would seem to me that if the housing is stationary and the axle measurements are indicating an off center condition wouldn't that be of the of the axle? Imagine the axle in the housing only supported by the inner Hyatt bearings and a perfectly ridged straight axle in a straight housing. If the axle was rotated I would expect that the axle end would not move. Now if one bearing was slightly off center and stationary what would cause a perfectly straight rigid axle to indicate a wobble as indicated in below when it was rotated? I would think that the pitch of the axle would be constant and no wobble.
https://mtfca.com/phpBB3/download/file.php?id=77036&t=1
When was the last time the axle was rebuilt? Like Hank Lee noted it could be something in the differential cause the axle shaft to ride on cause the movement. In any case, given the wobble, time to take things apart to determine the cause regardless of what it may appear to be.

[Bill Robinson]

Frank- I don't recall stating that there is any wobble to the axle shaft as it turns. Refer to my chart where I measured the shaft as it turns. There is no visible wobble. The little difference in the measurement at each quarter turn could either be 1. irregularities in the internals of the axle tube, or 2. lack of ability of the person in charge to use the internal measuring device.

[TRDxB2]

Frank- I don't recall stating that there is any wobble to the axle shaft as it turns. Refer to my chart where I measured the shaft as it turns. There is no visible wobble. The little difference in the measurement at each quarter turn could either be 1. irregularities in the internals of the axle tube, or 2. lack of ability of the person in charge to use the internal measuring device.

I have no doubts about your ability to measure accurately. I didn't say that you stated that there was a wobble. What I was referring to, given your measurements, that the axle would appear to wobble if it was rotating continuously (lower measurement from .647 to .650). You measured up and down do you have measurements from < side to side > ?

[Bill Dizer]

What was the problem that was being checked when the off center axle was noticed? Did the bearing slide out easily or did it take a lot of force? Does it slide back in easily? Bearing tolerances are fairly wide-wear of five thousandths is acceptable to most of us. Don't over think the problem, you are dealing with antique equipment and what worked fine in those days may seem unacceptable by modern standards!

[Charlie B in N.J.]

Going to try again. Doesn't matter if the other wheel is off or on or if it's on a jack stand off the floor. Both bearings are installed on that side. Anyone that has ever taken off an outer axle bearing has seen the shaft move off center of the housing hole. It's way more apparent if both outer bearings are out. The weight of the diff assembly causes that. Rotating the opposite wheel to rotate the diff assembly might cause the axle to describe an arc or it might stay in the same off center attitude and just turn. If it describes an arc (forms a circle if you will) it does not mean anything is bent. It just means the axle end is un supported and the diff weight is causing it to go off center. Put the bearing back and rotate the diff with the other wheel. If the axle doesn't swing in an arc it's not bent. So why or how did you come to the conclusion that the housing it bent?

[Allan]

In a perfectly straight axle housing, the axle will be below centre at the outer end. Gravity will pull the axle down and the inner bearing will not counteract this. Tolerances within the bearing and wear on the axle bearing journal will allow this 'sag' in the axle. The check I use is to lift the axle by hand. If it then appears to be the same distance from the top of the tube as it was before, then things are OK. A Herm suggested in the link, If the same measurement applies when the axle is pulled forward and then pushed backward, then things are OK on the horizontal plane too.

The worrying thing here is the ale is well ABOVE centre to start with, and the problem is likely to be worse when the shaft is lifted.

AllAn from down under.

[Pep C Strebeck]

Just a quick question: Have you used a straight edge against the outside of the axle housing/tube to see if it is bent, or how much?

[Original Smith]

If you have someone available who knows how to straighten housings, go for it. Most all of them are bent to some degree.

[Bill Robinson]

This morning John Z and I dis-assembled the axle. As stated earlier- it appears that the RH half of the axle housing may be slightly bent down, towards the road.
The 1st photo shows the female end of the axle tube where the axle bearing sleeve is inserted. It is extremely grooved with highs and lows from end to end. The photo shows the grooves AFTER we honed it the smooth the high points a bit. After honing for about 30 seconds, then inserting the sleeve, we both immediately noticed that the gap in the sleeve had spread some. We then inserted axle shaft and then put the Hyatt bearing onto the shaft and easily slid the bearing into the sleeve. Before the honing, a little tapping was required to insert the bearing. There was no play between the shaft and the bearing. It felt good.
Photo 2 shows a straight edge on the bottom side of the tube. There is a very slight amount of daylight visible regardless if you look at the top or bottom.
Photo 3 shows a defect in the top of the axle. It appears to be a weld that someone attempted to grind smooth, but missed their goal. It now appears possible that the axle may have been involved in a wreck at some time.

[Jerry VanOoteghem]

Some folks straighten axle tubes by welding on one side. The thought being, that the weld will cause the metal to shrink on that side, thereby pulling the tube straight. Not sure if the weld you show was done with that intent or not.

The rough bore in the tube is pretty much normal. It's a nice fit when the bearing requires a small shove, (usually by hand), when inserting it.

To get a real idea of how straight your tubes are, bolt the two halves back together, then set the assembly on 2 large "V" blocks. Rotate the axle housing on the "V" blocks and look for runout. These don't have to be precision "V" blocks. Anything you can make out of wood, that's solid and large enough will do. Place the blocks as close to the backing plates as possible, then rotate. Keep in mind, the mating faces of the differential housing can also be warped.

[Bill Robinson]

Great idea about the "V" blocks, Jerry. I don't know if it will tell where the "crooked " is, but it sure will tell me the "IF" it is crooked and where to start looking..

[TRDxB2]

Another measurement method found on MTFCA

[RajoRacer]

That's how I check mine - I use my c.c. jig that was ground as a surface plate.