[yngndrw]

Hi,

I'm wondering how you would take a block of metal which had 6 un-even sides and face each side so that opposite sides were parallel and adjacent sides were perpendicular.

I'm looking for a step-by-step guide to doing this accuratly, with details for how you would clamp the part.

This might seem like a very simple operation, but the thing that I don't get is how do you clamp the part to keep the next face parallel / perpendicular from a reference face.

The way that I can see to do it is:

* Clamp the part and machine the first reference face.
* Turn the part over so that this reference face is flat on the bed and machine the opposite face. (But how would you clamp it, making sure that the part is clamped enough that it doesn't move, but the reference face is still held flat to the table ?)
* Then I don't know how you'd go about it.

I've made a little picture to show the problem with part 2. (Black is the bed, grey is the vice and red is the part.) - Surly if you clamped the part enough, the reference face would lift off the bed ?

Thanks,
-Andrew.

[Geof]

Change your sequence.

Grip the uneven piece in the vise and face off the top surface.

Now put this faced surface against the fixed jaw of the vise, put a small block of metal (I use a nut) between the moving jaw and the uneven face of the piece so that the clamping pressure is centered in the fixed jaw; this holds your faced surface flat against the fixed jaw.

Face off the next surface.

Now repeat the clamping with a little block with the second faced surface against the fixed jaw and the first faced surface against the bottom of the vise.

Face off the third surface.

Now you have two faced surfaces parallel to each other and perpendicular to the other face within the accuracy of your vise. Gripping the piece is much easier now and facing the fourth surface is simple.

To do the ends align it vertically by using a dial gauge held in the spindle to run up and down the sides not held by the vise jaws.

[MrWild]

It helps to have something I call a knife parrallel. This is a parrallel that has one end with a wide base that is placed against a vise jaw, and the thin end (not really a knife edge) goes against the stock.

First step is to machine the top face flat. Once flat, this face goes to the rear and mates to the rear jaw of the vise. The knife parrallel is used as a go between the part and the moveable jaw of the vise.

Machine the exposed width. You now have the original top and a side square to each other. Flop the part so the original top is to the bottom and the machined edge is to the back. Use the knife parrallel so that the moveable jaw isn't changing the way the block sits when clamped. Some small parrallels betwen the visde and stock will let you tap the block with a l;ead hammer or dead blow until both front and rear parrallels are tight.

Now machine what was the bottom in your first cut.

Flip the stock so the machined edge from the second step is now on the bottom. Both mai9n faces are to the front and the back and the knife parrallel is no longer needed. Tap the block down onto parrallels once more so they are tight and machine the exposed edge.

With both faces and twoedges done, it is a simple matter to use a long enough endmill to machine the sides of the block hanging out from the edge of the vice.

The main thing is that the vise is trammed and square to the table, and the head is trammed and square also. If the head isn't trammed or the vice isn't square, the best you can do is still garbage.

[yngndrw]

Thanks guys, it makes sense now.

I'm wondering about this knife parallel - I couldn't find an image of one on Google, what exactly do they look like ? It seems to replace the scrap block / nut that Geof mentioned.

Thanks for your time.

[Geof]

....I'm wondering about this knife parallel .... It seems to replace the scrap block / nut that Geof mentioned.

Thanks for your time.

Yes, more or less. It gives a line contact, mine gives a point contact but the idea is the same to avoid the moving pushing the piece askew from the bad surface.

[yngndrw]

Ah I see, thanks for your help then.

[gsrmmeza]

Look up Starrett hold-downs. The images from Google are only showing the top view. If you have a Starrett catalog (printed version) it may have a better view. The way they work is after you face one side you put the set of hold downs into the vise. The machined surface goes against the bed of the vise and one hold down goes between the hard jaw and the part. The other goes between the movable jaw and the part. This will only get two sides parallel to each other. You would the clamp across the two parallel sides and then cut a third side. That gives you two square corners. Flip the part and continue with the rest.

[dertsap]

you can clamp it in the vise and side mill roughly 1/8 or 1/4" deep steps on 2 opposite sides ,
just enough to clamp those surfaces in the vise ,then the part will be securely clamped , you can face and sidemill 5 sides then flip it over and face the ther side

generally you'll want to leave a little extra so that you can finish all sides and to be sure to not have any warpage if your removing much material or if there is scale

[ImanCarrot]

All very good advice. The circular bar trick by Geoff is incredibly useful... here's a link to explain:

http://www.jjjtrain.com/vms/mill_squ...uaring_06.html

[yngndrw]

Ah I've just looked up a picture, makes more sense now.

I like the idea of face and sidemilling 5 faces in one go - I guess it's only for small parts though.

Thanks for the link, I'll take a look now.

Cheers all.

Edit: This link is a really great resource, thanks !

[Arquibaldo]

It's strange that nobody mentioned the old trick that allows one
to face the last 2 faces without using a square or an indicator on
the part and also not cutting with the side of the tool...
It is possible and is usually the first exercise in a milling course...
Hint: You must perform 3 cuts instead of only two for the last
2 faces.
If interested, I will post a link

Nelson

[Geof]

Take a skim cut with it as perpendicular as possible. This generates a surface that is perpendicular to the fixed jaw.

Now invert and rotate and use this surface to stand it straight, finish the other side and go back and finish th first.

[yngndrw]

It took me a while to work that one out, even with Geof's explination.

After the first of the three cuts, you end up with a "line" which is perpendicular. I.e. A face which is skewed in only one axis.

Cheers for that, may save some time.