Removing bearing race from magnesium upright

[stownsen914]

I’m replacing wheel bearings on a magnesium upright from a March 87C (used on a GT Porsche). I thought I had a plan to remove the bearing races from the upright, but realized it won’t work. If I heat the upright, will it be easy to remove the races by carefully tapping out with a punch? Or would it be preferable to have a mandrel of appropriate size to apply the force more evenly? I do have a shop press.

Here’s a pic showing the upright, and the race I need to replace. Thanks for the input!

[Rick Kirchner]

just did this myself. I use anti-seize on the bearings before putting them in and at 350 deg they wouldn't budge, but at 425 they came out easily.

I just used the oven in the kitchen.

If you don't have the appropriate tool, big sockets, various pieces of bar stock, etc usually works.

[stonebridge20]

NEVER remove or install a bearing race in an uprite that isn't heated first. I prefer a rosebud torch run evenly around the outside of the uprite when removing a race as it heats the uprite quicker than the race which will grow also with heat. I use a kerosene heater to heat the uprite evenly for installation of the race. I know I'm hot enough when spit bounces off the uprite.

[Phil Creighton]

I'm with Mike. If the casting is hot enough for spit to ball and bounce off its not going to expand any more. Was taught that trick a long time ago when I worked for a racing transmission co. in the UK. Bearings etc in freezer for re-assembly helps and you need to be quick. Mikes right - rosebud tip on oxy-acetylene is the quickest way, did an FGA CWP that way only yesterday.
Phil

[Rick Kirchner]

I fear I am insufficiently skilled with the torch....

Several years ago I ruined a "spare" (welded) upright in an experiment to see if molasses would clean it, and I should have kept the part for practice for situations like this.

[BorkRacing]

Torch or oven works because the coefficient of thermal expansion for steel is between 11 to 14, and for aluminum it is 23 i.e. approximately twice of steel. Uprights are not pure magnesium but an alloy of aluminum and magnesium but you get the idea...

[Phil Creighton]

The trick is to keep the torch moving constantly and keep coming back to the thicker areas of the casting. Put the casting on wood blocks rather than a metal table so it doesnt soak up the heat. Keep it moving and protect any seals that may be staying in - for example when doing backlash on a CWP and you are moving shims about. Find a weight to put on the freshly installed bearing or race till it cools off a bit and the bearing doesnt drop out. Try not to set fire to any rags/paper towels or the blocks of wood that are on the bench!!!!

[Rick Kirchner]

I set a rag on fire once. it was right between my feet, I was welding, and the weld spatter lit it off. Of course, the hood makes it hard to see much. Then there was the what's that light? What's that smell? Why am i so hot????

[billtebbutt]

Use a torch. Unless divorce is something you're keen on....

bt

[stownsen914]

I set a rag on fire once. it was right between my feet, I was welding, and the weld spatter lit it off. Of course, the hood makes it hard to see much. Then there was the what's that light? What's that smell? Why am i so hot????

Hey at least it was just a rag. I set my shirt on fire once while welding. Also didn’t notice until I smelled/felt it!

[stownsen914]

I don’t have an oxy torch. Would mapp gas suffice?

[BeerBudgetRacing]

Use a torch. Unless divorce is something you're keen on....

bt

I used the oven for a ring gear once. Once being the operative word.

I've used the BBQ for uprights before, but I agree about the expansion of the bearing race.
Torches are a must!

[ejsprint30]

I have to chime in cause I did this twice over the last week. Oven at 350-375 with upright and bearing in for an hour and a half, and bearing came right out. This on a cast aluminum radial upright for DP08.

Just wanted to ad that I have an industrial oven in my shop that I used to use when I had my cryo business to heat cycle after the process...so I did not stink up the house.

[Rick Kirchner]

my new shop is going to have an electric stove, and a washer/dryer just so I can do shop stuff without messing with the good stuff.

[Bob L.]

I bought a used oven from a Rebuilt store (30 bucks) mounted it on wheels and plug for welder outlet in shop, (spent more on those than for used oven...) Clist is also good source for old ovens/toaster ovens/air fryers, size is key. Once out hot, I use a HF heat gun to keep up with lost heat while in press. Worked well on all 4 corners of my RT5. 2nd pair of hands would have been nice.

Torch can work, but figured it would disturb the Keyphos coating I had applied to mag parts....not to mention tough to control heat.

I've also soldered radiator tanks/necks with the heat gun...slower heat, easier to control than torch. HF offered a small butane torch than is handy and easier to control, than cutting/welding/rosebud torch.......Oven makes everything even temp. to start

Don't forget you don't want to put out any mag. fires with water!

Bob L.

[Rick Iverson]

I know I'm hot enough when spit bounces off the uprite.

OK, this made me laugh

[stonebridge20]

OK, this made me laugh

Spitting at hot uprites is no joking matter Bro.

It takes skilz!

[johnhafkenschiel]

my new shop is going to have an electric stove, and a washer/dryer just so I can do shop stuff without messing with the good stuff.

Rick: A dishwasher is helpful for cleaning parts.

[Rick Kirchner]

thats a good idea, especially if you figure out how to inject a bunch of oil eater

[stownsen914]

I realize I should be aiming for the "spit bouncing off the upright" temp , but is there a numeric temp I should be aiming for?

[bob darcey]

Might be an answer here: https://en.wikipedia.org/wiki/Leidenfrost_effect

[Rick Kirchner]

When I put my bearings back in I heated the housing to 375 and the bearing was in the freezer, call it 25 degrees, so about 350 deg difference, and it dropped right in, could have used less heat.

When I removed them I had to go to 425, and of course the bearing was the same temp and it took a little on the press to pop them out.

It's something like a .002 press fit, I'm sure somewhere there's a formula for the growth of mag alloy vs steel that would give you a target temperature.

If I had to guess with the torch method, about 400.

[stownsen914]

On the recommendation of a shop I've worked with in the past, I got a propane torch setup to heat the upright. It's substantially larger than the Bernzomatic one I tried earlier.

Another question - for one of the bearing races, there is very little to grab onto to press it out - less than 1 mm around the perimeter. Suggestions on methods to remove after heating the upright?

[teamwisconsin]

Once it’s hot, use a long punch. Those three round little reliefs on the backside? Tap the punch on one, then move to the next. Keep going around. 2-3 good shots on each position and the race should come out. Just be cautious you don’t let the punch scrape the bearing bore as the race works it’s way out.

[stownsen914]

Once it’s hot, use a long punch. Those three round little reliefs on the backside? Tap the punch on one, then move to the next. Keep going around. 2-3 good shots on each position and the race should come out. Just be cautious you don’t let the punch scrape the bearing bore as the race works it’s way out.

Thanks. That's what I was thinking too, I just worried that it would be easy to get the bearing cocked in the bore and then I'd be hosed.

[Bob L.]

One other trick from "industrial world": with races that aren't going back into bores... run a bead on ID of cup/outer race.... let it cool and often the race will fall out.... (if worried about something to pull/push on, weld tabs/nuts to id). On cooling weld bead contracts/shrinks, pulling race inwards,...won't harm the the bore. Of course the old brg is junk... TIG torch is cleanest as less splatter than stick, but either will work.

Bob L.

[stownsen914]

Wound up getting a propane torch that runs off a grill tank. That was plenty to heat the upright to 350 degrees and the bearing shells came out pretty easily. Same temp + new shells in the freezer, and they dropped right in.

[Rick Iverson]

Gents;

So the rear upright order of assembly is something like this:

• Heat Mg upright to 425F (auto ignition is 883F);
• Copaslip anti-seize compound on the bearing OD surface (dissimilar metals);
• Dry ice for the stub axles;
• Heat brake rotor flange to 425F;
• No Loc-tite on the stub axles;

There are six (6) 1/4”-20 x 0.500” bolts securing the bearing in place.

• Torque?
• Loc-tite?
• If so, what color?
• Or safety wire in lieu of Loc-tite?

Since the bearing race is slightly taller than the hub, I will machine a ledge off the washer to provide flat mating surfaces.

Thoughts?

[tlracer]

To reduce corrosion between dissimilar metals, I have always used Duralac:

http://www.llewellyn-ryland.co.uk/downloads/duralac.pdf

I'm not sure whether it would be suitable in your particular application due to the heat involved in fitting the bearing, but may be useful info to someone, somewhere!

[stownsen914]

One observation was that I found it challenging to keep condensation from gathering and freezing on the chilled bearing shell before I could get it in the upright. I had coated it with oil before putting in the freezer, but the frozen condensation would appear faster than I could get it into the upright. A quick wipe with an oily rag helped a little, but some condensation had reformed on the shells when I dropped them into place. Is there a trick to doing that smarter? Or that's just how it goes?

[tlracer]

You're correct that a thin film of oil will reduce condensation forming, but I've generally figured it flashes off at the point where it gets close to the heated upright!

[ejsprint30]

I have always put the bearings in a zip lock bag with a heavy coat of WD 40 before freezing.

[Rick Iverson]

You're correct that a thin film of oil will reduce condensation forming, but I've generally figured it flashes off at the point where it gets close to the heated upright!

I have always put the bearings in a zip lock bag with a heavy coat of WD 40 before freezing.

Do you coat the heated hub with an anti-seize compound?

[problemchild]

Or that's just how it goes?

I have never done anything other than put the bare bearing into my freezer (which always needs a defrost). Never had a problem.

Between travelling in the trailers through all 4 seasons, my shop, and a barn, my cars are subjected to a full spectrum of temperature and atmospheric conditions. I don't consider a couple hours in a freezer as a concern.

[ejsprint30]

I have never used anti seize when installing bearings...just heat and freeze.

What I will say is that my car was an x pro car that obviously raced in the rain, and that was evidenced by rust on all of the bearings. I can't believe the difference now when spinning the tires

[DanW]

I have never used anti seize when installing bearings...just heat and freeze.

What I will say is that my car was an x pro car that obviously raced in the rain, and that was evidenced by rust on all of the bearings. I can't believe the difference now when spinning the tires

Completely agree. I thought it was odd to use anti-seize. Under no circumstances do you want the outer race to turn in the housing fit. You don't want the inner race to spin on the axle. The bearings need a light interference fit on the outer race and an interference fit on the axle.

The ABMA actually has a specification for bearing fits for various bearings and services called ABMA 7. SKF also publishes the fit specifications in their catalog.

[Rick Iverson]

I have never done anything other than put the bare bearing into my freezer (which always needs a defrost).

Completely agree. I thought it was odd to use anti-seize. Under no circumstances do you want the outer race to turn in the housing fit. You don't want the inner race to spin on the axle. The bearings need a light interference fit on the outer race and an interference fit on the axle.

The ABMA actually has a specification for bearing fits for various bearings and services called ABMA 7. SKF also publishes the fit specifications in their catalog.

Fellas;

Brilliant. Thank you.

[problemchild]

What I will say is that my car was an x pro car that obviously raced in the rain, and that was evidenced by rust on all of the bearings. I can't believe the difference now when spinning the tires

Many of the pro teams were and are doing "trick" stuff to wheel bearings that require frequent maintenance to keep them clean and adequately lubricated. The lack of ongoing maintenance was probably more of the issue than the rain itself.

[Raleigh]

It may be worth noting , Ralt service Bulletin's recommended spraying the outer bearing surface with silicone spray and placing in a zip lock plastic bag before the bearing is placed in the freezer when installing wheel bearings in Mag uprights.

One observation was that I found it challenging to keep condensation from gathering and freezing on the chilled bearing shell before I could get it in the upright. I had coated it with oil before putting in the freezer, but the frozen condensation would appear faster than I could get it into the upright. A quick wipe with an oily rag helped a little, but some condensation had reformed on the shells when I dropped them into place. Is there a trick to doing that smarter? Or that's just how it goes?