Milling a square lump of steel that isn't square

Wanted to make a rule holder so I can have the rule vertical on the surface plate for marking out. Just a clamp against a square block with a recess for the rule. Simple tool so simple job? - no!

 

It turns out (no pun) that the end of square bar (about 1.5" sq) I hacksawed off is not square in any direction. There is always an angle that causes it to lift between the jaws of a vice or when secured to the milling table. Have got one face fly cut but when attempting the next face it always lifts, no matter which face I try. There is always a slight angle or roundness which causes me problems.

 

So I am thinking that I might as well drill a hole through it so I can bolt it securely to the milling table, raise it using a spacer and then at least I can mill four sides square and will only have the top and bottom to do afterwards.

 

Can anybody's knowledge or experience come up with a better plan than this?

 

Thanks

 

John

 

Have you tried placing the block in the vice with the flat face against the non movable face and then clamping the block with a round bar trapped between the movable jaw and whatever the face looks like. That should hold the block to allow you to machine your second flat face at rt angles to the original reference. Having done that, repeat again which gives you three faces etc.

Kwil

 

That sounds brilliant sir. I shall give it a go tonight after work

 

Thank you

 

John

 

Hi John,

 

A couple of alternatives,

 

Use a 4 jaw as Keith says, you can either locate your good face against the chuck face, or reverse the jaws and locate against those or use a parallel to set the piece (take the parallel out before turning the face!).

 

On the milling machine use an angle plate to machine a face at 90 degrees to your original, set the second face vertical and mill a third face. Then use a parallel to set your second face to the bed and mill a fourth face and so on.

 

Set a piece of accurate flat or plate (preferably ground stock but bdms would do at a pinch) vertically in the vice, use this in lieu of an angle plate. use the top of the vice jaw as a datum if it is accurately ground (check with dial indicator).

 

As above but mill a face on the plate with a ledge to set the block on to to act as datum.

 

It all depends on the size of the block and the degree of accuracy needed. Personally in thiscase I think that the 4 jaw is the way to go. but it just goes to show that there are many ways to skin the proverbial cat.

 

Regards

 

Terry

As KWIL says the use of a soft rod against the movable jaw should stop lifting. Our friend Bogs gives a good description part way down this page.

 

J

In an ideal world one would use a hemisphere rather than a rod, giving a point contact rather than a line contact. But, of course, we don't live in an ideal world.

 

Regards,

 

Andrew

Could use a ball bearing I suppose.

I believe the recommendation is a half to two thirds of a large surplus, and softened, ball bearing, to avoid indenting the vice jaw.

 

Regards,

 

Andrew

Hi, Kwil as I was reading down these posts, and I was thinking the same thing, only I was going to say adding a suitable piece of flat with a good dimple in, between the ball and the moving vice jaw to stop the ball from pinging out unexpectedly.

Ninth video down - Squaring and Sizing Material - from http://littlemachineshop.com/video/videos.php

 

Hope this help, A

Hi Jason,

 

Using the round bar method is ok as long as there is a parallel line between the front and rear faces, otherwise one can find that there is only a point contact between the rod and the face of the work. Perhaps you are assuming that the faces are only tapered in one plane. A beginner may not realise that it is possible to attain only point contact. Even God didn't get it right all the time.  call me over cautious but I would still prefer an angle plate and two clamps or similar.

 

Hi Andrew,

 

surely a ball will only give a point contact and will allow rotation of the workpiece if something goes awry, no matter how tightly it is clamped It is not something I would attempt. lightly.  It is not a method any of my instructors would have allowed.

 

I would prefer the 4 jaw chuck method as Keith described, accurate and safe, but what do I know?

 

Regards

 

Terry

We may be overlooking the obvious.

 

The nearest approximation to square that we can get is a single point inscribing a disc at right angles to an axis. So you take a bit of thickish bar, clock it true in a 4 jaw and then skim the front face and relieve said face til a thin ring on the circumference remains. (That is important) Best is if it is infinitely thin, but the practicalities dont permit that.

 

That will be square- square enough to check commercially made squares in fact.

 

Attach rule as required.

 

Beyond that, I wonder we aren't making a simple thing difficult, finding errors when we know some will exist and then being very surprised, since nothing (other than a disc inscribed from a single point pretty much ) is truly square. There is always an error. However, the soft bar and the use of a machined face against a reference face is both time honoured and effective.

 

G'Day Keith Long and Keithmart,

Luxury!!! When I started as an apprentice we were given a lump of cast iron to chisel , then file, then scrape then polish into a shiny square block!!! Then again it was in Australia

Regards

Dave Tointon

 

Even without a 4 jaw chuck, a chunk of squarish steel can be clamped in a 3 jaw, to face it, it does not have to be centered, just square it against the chuck face, I,v done a few bits that way, faced off bits of plate, and squared the edges, Its some times easier and quicker than using the milling machine. Ian S C

I described how to square up an approximately square chunk of metal a few years back with my articles on a dovetailed novelty. That was done with a shaper but the same process can be applied to a mill. First you select a good face to start with, the flattest one you have. Put that against the fixed jaw of the vice, with a small diameter round bar lying across the ways of the vice for the workpiece to lie on. Put a piece of thick cardboard, then wood, or customwood against the moving jaw and bring it up against the job and tighten it. The idea is that the relatively soft material will crush in a little and hold the good face against the fixed jaw of the vice. The small round bar under the job prevents that face from trying to alter the angle of the job.

 

OK, so provided the vice is square this setup will let you machine the top face so that it is at right angles to the face held against the vice. (If the vice is not square it needs attention, I had to adjust one of mine once.) Next step is to rotate the job so that the face we just machines is against the vice jaw and the original best face is downwards. If that face was really good, we can now dispense with the round bar, but we still use the soft packing on the moving jaw. We machine the new top face , and now we have two faces which we have machined and which should be at right angles to each other. The second one should also be parallel to our original best face, and next time we rotate the job that original face will be against the moving jaw, with the first face we machined lying on the ways of the vice. Because of this job should now sit very nicely in the vice, and should not need so much soft packing on the moving jaw side, although if the original best face was not too wonderful a piece of card might still pay. Next cut brings us to three good faces, and after that if the original best still needs cleaning up a little we can turn one more time and do that. We should at each stage make sure the job is tapped down against either the ways or the piece of bar, and in the early stages with the thick packing distributing the load we should make sure the cutting loads are not too excessive and are towards the fixed jaw.

 

OK, now we have a nice true rectangular cross section, but the ends are still not square. We also have to get it set up to be square to two sides, so we can't just use the same trick of ensuring that it is square against the fixed jaw. There are several ways to proceed, which is best will depend on the proportions desired, eg cubical or a longer rectangular shape, and also on the machinery.

1 Hold the 4 good sides in the 4 jaw and turn the two ends on the lathe.

2 Hold in the machine vice by the good sides and mill the end with an end mill or slot drill, with the end sticking out the side of the vice. The vice must set true on the table, and the cutter must be long enough to reach the depth...long cutters will flex so this will not give a super accurate end.

3 Put a V block against the fixed jaw in the vice and hold the job in the V block by two of the good faces with the moving jaw against an edge. This requires a V block that is not too big or two small for the particular size of cuboid. It does work quite well, although you may want some packing to stop the vice flattening the edge where the pressure goes on

4 Set the block in the Vice and use a square to line up the other face against the table or the jaw. This is not all that easy to do accurately.

 

It is not all that easy, even with good gear, to mill (or shape) a really true cube, and mgj is right about using a turned cylinder where a really accurate square is needed, although I would turn the whole thing between centres. If you mike the diameter at each end they should come out the same, if they don't your centres are not aligned. Which is another fun thing to spend a bit of time correcting!

 

regards

John

Graham - you are right- one CAN do everything with a file,and it takes a great deal of skill & practise. (I don't have either the skill or patience!) I have seen some of this apprentice work, and a marvle a lot of it is.

 

Sadly some would say? I don't, personally - , a lot of the need for it has been repalced by hte machine tool when it comes to munching chunks. (Tin bashers are a world apaprt). If I want a plane surface I wil generate one, by milling, flycutting or wahtever, but I accept there are errors - I don't beleive in absolutely flat and truly smooth or absolutely round! i don't beleive in CNC machinery either - not for my sort of work, not because it doesn't have huge advantages, but because its not me - I like handwheels and dials. I am not interested in a computer coming between me and what I want to make.

 

Back to this point -

I just felt that - while the bit about how to mill a square chunk was interesting, - the original question both overlooked the obvious, and over-engineered with less accuracy! (For that particular function. )

Hi mgj,

 

I agree entirely, Personally I think that there is over emphasis on extreme accuracy in the hobby fraternity. When I was industry we had levels of standards. i.e. from High to lesser it was, Inspection, Toolroom and Workshop. The latter was for the actual manufacture of products. When I see hobby equipment advertised as 'Toolroom Standard' I have to ask 'why', unless it is for the Experimental Engineer.

 

In this thread we are discussing a holder for a rule. If this is the only purpose of the block it just needs two faces at right angles. The rest don't really matter, then you have to consider how accurate do those two faces really have to be? If they are not exactly at right angles it won't affect the tool significantly. Just get on and do the job and don't worry too much.

 

Regards

 

Terry