Welded construction for surface grinder arbor?

[Flynth]

I've been intending to make a couple of additional surface grinder arbors for years as my surface grinder came with just one and if I take a wheel off it requires rebalancing.

The arbors have a 90mm diameter (almost 4in) collar on the back. This collar in addition to providing something for the wheel paper to push against have a face grooved dovetail and 3 weights with grub screws to balance the wheel.

The problem is that to make the arbors from single chunk of steel I would have to start with a 90mm+ diameter bar(almost 4in) just to turn 90% of it to around 25mm(1in).

So I have an idea to tig(or mig) weld 30mm diameter(~1 and quarter in) bar to the 95mm bar and machine it from there.

I thought to ask here first if this is a bad idea. There is a taper on the inside of the arbor so I would have to ensure the weld is as thick as the resulting material on the finished part. I can chamfer the smaller bar a lot and do multiple passes to fill it with steel.

I plan to use standard mild steel, and a standard mild steel welding rod/wire. I plan to wire wheel clean it between passes.


What do you think? Is welding it worth it? This arbor has to be super precise. Am I likely to have some internal stresses in the part that may reveal themselves when I cut the taper near the end? Is mild steel a bad material choice perhaps? I don't have any case hardening compound available to me so it will stay soft.

 

[projectnut]

I'm not sure what brand grinder you have or the size of the "arbor". I'm thinking by your description you're referring to what's more commonly known as the wheel hub. There are currently a dozen listings on eBay for wheel "hubs", and several more using the terminology "adaptor" or "arbor". One listing offers 5 hubs, and a few well used wheels for $125.00. The seller doesn't give much of a description as to the size, taper, or thread direction, but if they fit your machine, you probably couldn't buy the material for that price.

Keep in mind if your wheel rotates clockwise, you'll need hubs with left hand threads. If it rotates counterclockwise, you'' need right hand threads.

 

[Flynth]

I'm not sure what brand grinder you have or the size of the "arbor". I'm thinking by your description you're referring to what's more commonly known as the wheel hub. There are currently a dozen listings on eBay for wheel "hubs", and several more using the terminology "adaptor" or "arbor". One listing offers 5 hubs, and a few well used wheels for $125.00. The seller doesn't give much of a description as to the size, taper, or thread direction, but if they fit your machine, you probably couldn't buy the material for that price.

Keep in mind if your wheel rotates clockwise, you'll need hubs with left hand threads. If it rotates counterclockwise, you'' need right hand threads.

I always thought the right name was wheel arbor, but you're right, more precisely it is a wheel hub. However I misremembered the dimensions. The diameter of the part grinding wheel mounts is 5/100 of a mm under 51mm (a little over 2in). So I would be welding a 55mm bar not a 30mm one.

The grinder is a Polish Jotes SPC20A made in 1960s. I never saw such wheel hubs for sale anywhere. There are two "old equipment" suppliers here in PL I know. They specialise in parts for such old machines. Unfortunately they target customers that use those machines to make money. Their prices are so high it is almost always cheaper to make the part. The only bits I bought from them were oil windows and other tiny parts. They wanted quite a bit for them too.

For wheel hubs. If I remember correctly, few years back I was quoted around $350~$400 a piece depending on how many I want. Just to put things into perspective. They sell a replacement spindle for this machine for $3500 (it uses plain bearings so no expensive roller bearings are present), they wanted $1k for a table feed clutch (multi plate, two part, solenoid actuated wet clutch) for my old German mill. Etc. Occasionally one can find parts like this sold by individual people (usually refurbished), but I never saw wheel hubs. This, and the price makes me think that supplier wanted to manufacture it just for me. BTW, I paid about $1k for the machine when I bought it. It is a 1.5t machine, with an electric magnetic clutch, hydraulic feeds on X and Y axes (both continuous, and stepped on Y). It's a really nice machine. The only pain is a lack of more wheel hubs.

That's why I would like to make at least 2

I wonder if anyone made anything like it with welded joints. The important bit is that the back face, the bit the wheel goes on and the taper all have to be very precise. (down to few 1/100 of a mm, or a thou).

 

[Shotgun]

I had the same issue making a back plate for a 10" chuck. My lathe has a 2-1/8"x8tpi spindle. Very odd size, and not readily available. I used the solution you're proposing.
A) make it oversize so that you can turn it to the correct dimensions.
B) heat the parts in a toaster oven to the highest temp available before welding. That won't eliminate all stress, but it will improve the situation.
C) use mild steel filler, just like you propose. The mild steel should be weaker than the parent material, allowing it to stretch more as it cools.
D) "ring" it as it is cooling. Tap on it with a hammer to introduce vibrations, that will work to help the material settle instead of locking in the stresses. I don't know how much this does, but my uncle taught me to do it, and it sounds feasible.
E) put it back in the oven to let it cool very slowly.

Since you're going to have lots of extra material, take off half the extra all the way around, then wait till the next day to finish machining it.

 

[jwmelvin]

I think you’d be better off using a shrink fit or retaining compound, to avoid distortion. Maybe brazing, but welding seems like an invitation for difficulty. If you leave sufficient margin to machine true and stress relieve, then I suppose it could work. I just don’t see the advantage over interference or adhesive.

 

[Flynth]

I had the same issue making a back plate for a 10" chuck. My lathe has a 2-1/8"x8tpi spindle. Very odd size, and not readily available. I used the solution you're proposing.
A) make it oversize so that you can turn it to the correct dimensions.
B) heat the parts in a toaster oven to the highest temp available before welding. That won't eliminate all stress, but it will improve the situation.
C) use mild steel filler, just like you propose. The mild steel should be weaker than the parent material, allowing it to stretch more as it cools.
D) "ring" it as it is cooling. Tap on it with a hammer to introduce vibrations, that will work to help the material settle instead of locking in the stresses. I don't know how much this does, but my uncle taught me to do it, and it sounds feasible.
E) put it back in the oven to let it cool very slowly.

Since you're going to have lots of extra material, take off half the extra all the way around, then wait till the next day to finish machining it.

Thank you. That sounds like a good method. I have an electric oven I could heat it in.

I think you’d be better off using a shrink fit or retaining compound, to avoid distortion. Maybe brazing, but welding seems like an invitation for difficulty. If you leave sufficient margin to machine true and stress relieve, then I suppose it could work. I just don’t see the advantage over interference or adhesive.

I suppose that could be done. A step would have to be machined so the larger part has no chance of flying off the end. It is an interesting idea I haven't thought about. Perhaps it is a bit more complicated than just welding though. I definitely wouldn't trust a braze there.

The plan is to have at least 5mm extra steel after welding. As for stress relieving. I wonder, does anyone here know the proper temperature and time? (I have a self built hardening oven and a gas forge).

 

[Liljoebrshooter]

Thread them together and do the finish machining.
Joe

 

[projectnut]

I think you’d be better off using a shrink fit or retaining compound, to avoid distortion. Maybe brazing, but welding seems like an invitation for difficulty. If you leave sufficient margin to machine true and stress relieve, then I suppose it could work. I just don’t see the advantage over interference or adhesive.

I think jwmelvin has the best idea so far. Over the years I've been involved with both welded and shrink fit assemblies intended for precision work. While the welding can be done it always involves extensive machining afterwards to eliminate warpage and insure stability. Shrink fitting allows you to machine the parts to press fit tolerances. Then they can be cooled to the needed temperature and assembled. Once assembled they can be pinned to assure they never come apart.

 

[Flynth]

Thread them together and do the finish machining.
Joe

Thank you for an idea, but it seems this would mean extra work over shrink fit with little benefit. I would still need a pin to prevent it unthreading.

I'm seriously considering the shrrink fit idea.

I think jwmelvin has the best idea so far. Over the years I've been involved with both welded and shrink fit assemblies intended for precision work. While the welding can be done it always involves extensive machining afterwards to eliminate warpage and insure stability. Shrink fitting allows you to machine the parts to press fit tolerances. Then they can be cooled to the needed temperature and assembled. Once assembled they can be pinned to assure they never come apart.

Yes, pinning would definitely be required. I'm just wondering how to do it. The 90mm diameter is 19mm long with only 3mm available due to that back face groove for the balancing weights(the groove is 16mm deep). Traditionally one drills across both parts to pin them. Here with just 3mm to work with that isn't possible. I guess I could make a V groove where both parts mate and flow silicon bronze braze in there. I wouldn't trust a braze for all of the force, but to retain sharing fit parts seems like a good idea. However, it would be something that is done almost as the last thing. If I then overheated and warped thd part game over, start again.

I'll add some photos later in case my explanation isn making it clear what I'm trying to do.

 

[Flynth]

Here are some photos of the original.