Grinder wheel axle flex/wheel alignment

[Hubert S.]

What approximations were you making about the stiffness of the wheel and where it was supporting the cantilevered beam? I just measured the 2" idler wheels on my D-plate and found they deflect about 2.5 thou at the far end under belt tension. I could throw my SGA with 5" wheel on to see what it does in a bit; I'd expect it to deflect even less because I'm using a shorter (and steel) spacer for it.

I did not make any assumptions about the wheel, just checked the maximum deflection for a round piece of steel with a point load to see if it is in the ballpark. I used a 3" steel beam (E=30Mpsi) with a 50 lbf point load at 2" and a diameter of 27/64" and got about 5 thou max. deflection. The load is just a guess and probably varies a lot between different grinders, but 50 pounds seems reasonable to me. There are probably other sources of flex, but it is plausible to me that a 1/2" bolt could flex a few thou on its own.

I mentioned earlier that I ordered a wheel for the SGA I am building with 20mm ID bearings. Because the deflection is inverse proportional to the fourth power of the shaft diameter, this setup should be about 12 times stiffer than using a 1/2" diameter bolt, and six times stiffer than using a solid 1/2" shaft.

 

[Natlek]

The bolt (axle) flexes.

1/2" bolt , tightened on 7-8 kg-m / 60-70 N-m and two bearing and spacer between head of bolt and tool arm, can flex just from belt tension ? Are you sure there is no play in bearings ? C3 , C4 bearing have some noticeable play when cold ? I need to check this on my own , hard to believe this , guys .

 

[Natlek]

with a point load

Hubert , what is point load ? This on picture ? If that is point load , the force from the belt is transmitted to the entire length of the bolt outside of tool arm ?

1/2" bolt could flex a few thou on its own.

But bolt is not on its own in our case ?

 

[Hubert S.]

Hubert , what is point load ? This on picture ? If that is point load , the force from the belt is transmitted to the entire length of the bolt outside of tool arm ?



But bolt is not on its own in our case ?

Natlek, I just did this to get an order of magnitude. The term "point load" just means that all the force acts on a single point. The point load was not at the end of the bolt, but roughly in the middle of the belt. Using a distributed load changes the deflection slightly, but not by an order of magnitude. I just looked at a bolt by itself, if you have a sleeve between the bearings and crank down on it, I suspect that you might gain some additional stiffness and lower deflections.

 

[Hubert S.]

1/2" bolt , tightened on 7-8 kg-m / 60-70 N-m and two bearing and spacer between head of bolt and tool arm, can flex just from belt tension ? Are you sure there is no play in bearings ? C3 , C4 bearing have some noticeable play when cold ? I need to check this on my own , hard to believe this , guys .

All I can tell you is that the cantilever beam equations for a bolt produce roughly what Josh reported for the assumptions I made. I have not measured anything. If you do check this on your own, can you do two tests, one where the indicator is mounted to the bench and one where it is mounted to the tool arm? This would tell you the total deflection and the deflection of the wheel relative to the tool arm.

 

[Natlek]

All I can tell you is that the cantilever beam equations for a bolt produce roughly what Josh reported for the assumptions I made. I have not measured anything. If you do check this on your own, can you do two tests, one where the indicator is mounted to the bench and one where it is mounted to the tool arm? This would tell you the total deflection and the deflection of the wheel relative to the tool arm.

It's too cold tonight to go there and try. Tomorrow during the day I'll do it.I hope you don t use some cheap mild steel or whatever bolts ?? 10.9 is what I use for that ?That make some difference also ? If i don t forget i will take with me from my service tool for measurement belt tension so i can try diff. tension

 

[CallumRD1]

It's too cold tonight to go there and try. Tomorrow during the day I'll do it.I hope you don t use some cheap mild steel or whatever bolts ?? 10.9 is what I use for that ?That make some difference also ? If i don t forget i will take with me from my service tool for measurement belt tension so i can try diff. tension

The grade of the bolt won't make a difference because the stiffness of a piece of steel (Young's modulus) is largely independent of variations in steel composition or hardness. This assumes you stay below the yield point of the material which we absolutely are in this example.

 

[Drew Riley]

EVERYTHING flexes. Some things just flex more than others. Start adding in multiple parts (platen frame, tool arm, grinder frame, etc... and you're bound to have at least a little flex somewhere.

 

[Natlek]

EVERYTHING flexes. Some things just flex more than others. Start adding in multiple parts (platen frame, tool arm, grinder frame, etc... and you're bound to have at least a little flex somewhere.

Yes , but M12 bolt under belt tension ?

 

[REK Knives]

All I know guys is that it flexes even w/ everything locked down tight. I'm not sure exactly where it's coming from but I suspect it's probably coming a little from everything. I can take a video w/ an indicator if anyone would like to see how I'm going about it?

 

[Natlek]

The grade of the bolt won't make a difference because the stiffness of a piece of steel (Young's modulus) is largely independent of variations in steel composition or hardness. This assumes you stay below the yield point of the material which we absolutely are in this example.

It make a difference with tightening force we can apply on bolt .

 

[Natlek]

EVERYTHING flexes. Some things just flex more than others. Start adding in multiple parts (platen frame, tool arm, grinder frame, etc... and you're bound to have at least a little flex somewhere.

Yes , everything flex if we apply required force .
I will waste my time but i will make some test tonight !
Drew , where would be flex in our case ? Can you show me on this drawing . If bolt is tightened properly what will flex ? Bolt ? HOW and where ? All parts are tightly tightened together so to get wheel to flex , spacer between tool arm and wheel has to raise on one side . I can not see any other way to get flex in that setup ? I f I explain right what I mean ?

 

[Natlek]

I am pretty sure every grinder uses bolts to secure the wheels to the tool arm/platen and bolts are precision ground. I recommend checking multiple angles as well (not just from one side). Now given, this is a 2.5" wide wheel and this is the far side of the wheel but it's a .005" gap (checked w/ feeler gauges).

*update: axle flex is apparently a significant issue for any of the wheels

When I rotate the axel/bolt around I get this:

Fortunately I was able to use an indicator and bend the bolt straight so it's good to go now! I am thinking about trying a step bolt as those are likely more precise.

Read this , pay attention on post 4 and 7

How to mount contact wheel on belt grinder

https://bladeforums.com/threads/14”-contact-wheel.1768294/page-2#post-20248865 I don t know if it is my bad English or what it is it , but in that thread somehow I turned out bad ................ I drew this drawing to explain once and for all how to mount wheel properly on grinder . And to be...

bladeforums.com

 

[REK Knives]

Read this , pay attention on post 4 and 7

How to mount contact wheel on belt grinder

https://bladeforums.com/threads/14”-contact-wheel.1768294/page-2#post-20248865 I don t know if it is my bad English or what it is it , but in that thread somehow I turned out bad ................ I drew this drawing to explain once and for all how to mount wheel properly on grinder . And to be...

bladeforums.com

I have 2 things that we are discussing...
1. wheel alignment
2. wheel flex

Yes I just read your posts and I do agree with what you are saying. My set up is like #2 in your OP of that thread. However, when I rotate the axle/bolt, the light gap DOES change, I'm not sure why. When I only rotate the bushings (between the wheel and tool arm) it does not change. Maybe I can try rotating just the bushing between the bearing and see what happens?

Regarding wheel flex, everything is tight and I'm getting about .006-.007" flex at the outside rim of the wheel. I've also checked my drive wheel for flex... my 5/8" arbor was worse than the motor I just upgraded to (which has a 7/8" arbor) but it still flexes about .003" or so.

 

[Keith Nix]

It's just not accurate for anyone to say there's zero flex in a half inch screw. ALL materials of all sizes deflect under load.
I think the question here is not whether there's deflection, but to what degree it actually matters.
When I dusted in the chuck on my SGA, I honestly didn't even think of that. So my chuck is ground parallel to whatever deflection is inherent in my loaded grinder.

 

[REK Knives]

It's just not accurate for anyone to say there's zero flex in a half inch screw. ALL materials of all sizes deflect under load.
I think the question here is not whether there's deflection, but to what degree it actually matters.
When I dusted in the chuck on my SGA, I honestly didn't even think of that. So my chuck is ground parallel to whatever deflection is inherent in my loaded grinder.

exactly, I agree 100%

 

[CallumRD1]

I always have the same belt tension on my SGA (I have tool-arm stick out marked for repeatability) so the few thou of deflection in my wheel is mapped onto my chuck and I can grind very precisely if I'm careful but this raises an interesting question: when grinding aggressively the belt is being pinched between the wheel and your work piece with the motor pulling the belt through with a very large amount of force, far greater than the static belt tension. This will cause further deflection and pull the wheel out of square with the chuck. I'm fairly sure this is why my SGA always removes more material from the work piece at the bottom of the wheel than the top when hogging. As the cuts get lighter this evens out and thus isn't an issue but I had never made the connection before. I'm currently experimenting with a few solutions but it will be a little while before I'm ready to share the results. I turned a stainless steel 20mm shaft with a flared base to use to mount my wheel. It will be screwed into the tool arm from below and a wheel with 6204 bearings will be a very tight slip fit or very light press fit onto it, then clamped with a screw from above. This should make the wheel more than an order of magnitude more rigid and decouples the two jobs the 1/2" bolt does in a normal assembly: preloading the bearings and securing the assembly to the tool arm. I'm currently waiting for bearings to arrive and I need to turn another wheel to fit them.

 

[Hubert S.]

I just measured a 10" contact wheel on my grinder with the indicator mounted on the grinder frame, not the tool arm (not enough room). I get about 5-7 thou deflection with regular tension and a bit over 10 thou if I really push on the tension lever. This wheel is mounted the same as #2 in Natlek's sketch.

 

[REK Knives]

I always have the same belt tension on my SGA (I have tool-arm stick out marked for repeatability) so the few thou of deflection in my wheel is mapped onto my chuck and I can grind very precisely if I'm careful but this raises an interesting question: when grinding aggressively the belt is being pinched between the wheel and your work piece with the motor pulling the belt through with a very large amount of force, far greater than the static belt tension. This will cause further deflection and pull the wheel out of square with the chuck. I'm fairly sure this is why my SGA always removes more material from the work piece at the bottom of the wheel than the top when hogging. As the cuts get lighter this evens out and thus isn't an issue but I had never made the connection before. I'm currently experimenting with a few solutions but it will be a little while before I'm ready to share the results. I turned a stainless steel 20mm shaft with a flared base to use to mount my wheel. It will be screwed into the tool arm from below and a wheel with 6204 bearings will be a very tight slip fit or very light press fit onto it, then clamped with a screw from above. This should make the wheel more than an order of magnitude more rigid and decouples the two jobs the 1/2" bolt does in a normal assembly: preloading the bearings and securing the assembly to the tool arm. I'm currently waiting for bearings to arrive and I need to turn another wheel to fit them.

wow that's sweet, please keep us posted on results and how well it works!

I just measured a 10" contact wheel on my grinder with the indicator mounted on the grinder frame, not the tool arm (not enough room). I get about 5-7 thou deflection with regular tension and a bit over 10 thou if I really push on the tension lever. This wheel is mounted the same as #2 in Natlek's sketch.

yep, I mounted my indicator to the frame as well. that sounds about right.

 

[Hubert S.]

I always have the same belt tension on my SGA (I have tool-arm stick out marked for repeatability) so the few thou of deflection in my wheel is mapped onto my chuck and I can grind very precisely if I'm careful but this raises an interesting question: when grinding aggressively the belt is being pinched between the wheel and your work piece with the motor pulling the belt through with a very large amount of force, far greater than the static belt tension. This will cause further deflection and pull the wheel out of square with the chuck. I'm fairly sure this is why my SGA always removes more material from the work piece at the bottom of the wheel than the top when hogging. As the cuts get lighter this evens out and thus isn't an issue but I had never made the connection before. I'm currently experimenting with a few solutions but it will be a little while before I'm ready to share the results. I turned a stainless steel 20mm shaft with a flared base to use to mount my wheel. It will be screwed into the tool arm from below and a wheel with 6204 bearings will be a very tight slip fit or very light press fit onto it, then clamped with a screw from above. This should make the wheel more than an order of magnitude more rigid and decouples the two jobs the 1/2" bolt does in a normal assembly: preloading the bearings and securing the assembly to the tool arm. I'm currently waiting for bearings to arrive and I need to turn another wheel to fit them.

This is exactly what I plan to do on the SGA I'm building. I ordered a wheel with 6204 bearings.

What material are you using to turn your wheel?