Thread: Elliot 8-18 Surface Grinder

Hi Phil, my thought process went along the lines of this: Since my surface plate is flat (well flat enough) and if I achieve a zero deviation on the DI from one end to the other, if I then replace that surface plate with an equally flat magnetic chuck and set it up to be equally as parallel as the plate, don't I then have a machine setup that would grind a "perfectly" flat surface? WRT the chuck preparation, I assume it's important to keep the same side of the chuck facing the same side of the table? What does the grease on the table do prior to mounting the chuck? Is that to help prevent oxidation under the chuck? Cheers, Simon

I'm not getting it either.

We are talking about the ways on the table and the machine being the same curve?

Whats wrong with this "picture"?

More thinking

Stuart

p.s. I see I've been slow typing again lol

Originally Posted by simonl

Hi Phil, my thought process went along the lines of this: Since my surface plate is flat (well flat enough) and if I achieve a zero deviation on the DI from one end to the other, if I then replace that surface plate with an equally flat magnetic chuck and set it up to be equally as parallel as the plate, don't I then have a machine setup that would grind a "perfectly" flat surface?

All I'm trying to tell you, is that test isn't really valid. There seems to be some confusion between Flat, Parralel, or Straight here.

Originally Posted by Stustoys

Whats wrong with this "picture"?

The person in the picture does not have anyone to hop on the other end of the see-saw?

So I fixed it...

pt.png

Ok I'm going round and round in circles.(or is that curves?)

The best I can come up with using the "pirateship" analogy.
Isn't what Simon has done with the plate the same as putting a guy at each end of the boat with a string tight between them(thats your straight line). Then measuring the distance from the string to the pivot at BDC as the boat moves. The only way that distance can be the same* though one "swing" is if the boat travels in the straight line.............I think........... maybe.....what am I missing?.. lol


Stuart



*assuming the distance from each end of the string to the pivot is the same.

I think the confusion is that Phils boat is moving in a perfect arc whereas Simons 'boat' is not attached to the pivot point and has the option to move in a not so perfect arc. Simons arc depends on the different amounts of wear on the rails supporting his 'boat'.
But then, I am sure there is something I am missing as well.
Steam engines are so much easier.

Phil

OK. My head is starting to hurt. Phil, I'm going to sit in the corner and have a think.

Simon

I tried to give up, really I did.

I cant see how a flat plane can move past a fixed point in space at a fixed distance unless it is traveling in a (on a?) flat plane.

Well ok I can but it relies on the table and base having some funky curve in there ways* so that as the table moves to the right from center point and falls**, it tips to the right and V V......... that the chances of coming out within 0.003mm would be pretty low I would have thought.

Phil whats the answer...... I need more note pads lol

Stuart


*the reverse of the way I think they would wear. (not that that means squat)
**if it rises moving away from center as I would expect I cant get it to work at all

Hi Stuart,

I understand what you are saying. There is a chance that the ways have worm at just the right geometry so that the movement of the table describes a locus ( I think it's called) so that it is always equidistant from the plate to the spindle. However, I'm thinking this condition can only be met when the spindle is a specific distance to the table, an eigenvalue if you will. It can't satisfy the same locus at every height value from spindle to table so moving the spindle either up or down from the chosen spot I did the test should yield a different result. If the results are the same, what then?

Can I also point out that it is in my nature to ask questions especially when dealing with systems I am in a steep learning curve with. While I have learnt a lot in the last 6 years, it merely gives me more of an insight into how much I don't know. I don't want this to be construed as me arguing the point. Indeed I have nowhere near the knowledge or experience to start arguing with you guys let alone with people like Phil who do this stuff for a living.

Anyway, I got impatient and decided to make sparks. I put the wheel on that it came with and dressed it. I have no idea what grade the wheel is but it's grey in colour and quite "smooth"

The piece of steel is of unknown grade but I don't think it's plain mild steel. It was rather rough to begin with so I started with 2-3 thou grind moving about 40 thou with each pass. The final grind was about 1-2 tenths, probably more like 2 tenths as I started with a tenth but looked like I was doing nothing! All done manually and no coolant. The cuck is yet to be dressed.

Close examination shows very fine ripples but I'm not sure if they are to be expected for the machine, wheel combination and no coolant. I'm pretty happy for a first try. Anyway.

Pic to follow:

Simon

Difficult to get a realistic pic.



When I took the spindle apart and cleaned the bearings I thought they looked close to the end of their life but figured id clean and repack them to see what the results would be like.

Im surprised at the finish but what would I know? It could be a terrible finish for all I know!

Simon

Finish looks good, when you get the hydraulics running and have a bit of a play with different downfeed rates, you will start to get a feel for what works. Coolant helps. and spark out for a bit longer. ( easier when you don't have to wind the handle )

Ray

Simon, you may also be travelling the table a bit fast. I find on mine that I can change the finish significantly with table speed.
I don't mean the difference between very fast and dead slow, I mean between a very slight reduction inspeed from 'normal' - whatever that is. Seems to me to be a question of how much of the entire circumference of the wheel 'dwells' in each spot.
At 3000 rpm (my 'normal' spindle speed) my wheel with 22" circumference travels at 91 feet per second in relation to a stationary workpiece. So for each spot to be contacted by the entire circumference takes a 'dwell' of 20 milliseconds. So moving the table at 1"/sec would give a ripple (from microscopic deviation from round) of 0.02" or 0.5mm..... any faster will spread those wave patterns out further.
Because each spot is contacted by the wheel multiple times (forwards and back), those ripples get evened out at random of course. Sparking out eventually removes any remaining high spots in theory at least.
I'm guessing that a combination of step-over and spark out passes along with traversing speed will let you find the 'sweet spot' of you grinder for that particular wheel eventually.
The three new wheels for you are still safe here. They are/were white Norton and I think either 60 grit or a little coarser.

Do you always grind heavy things in bare feet?

Nows the time to used that granite, run a clock over it,

More impotantly is the famous 4 corner test. (Which requires 5 elements). Put something thats remotely the same size on each corner, over a pole and one in the middle. Then grind them. In a perfect would they should be the same. Not worth doing until you have dusted the chuck.

Originally Posted by Stustoys

Phil whats the answer...... I need more note pads loll

Do we agree ways wear? Lets say the ways are worn - minus 5 thou per side, so as it moves left to right, the table droops then rises 5 thou as it moves, left to right. as in comes through centre.

No problem, lets just regrind the chuck, even than we know the table is sagging plus / minus 5 thou. The chuck will clean up. Put a clock on or about the position of the wheel, that has just manufactured that surfurce, what are the chances it will be zero -zero, even thou we know the table has a sag?

I cant explain it any better. A master is something you check from, not too.