Holding material (general)                Author – Derek Mackenzie

            I have mounted two 3” capacity vices on the milling table, carefully checked so that their Y position is identical (checked electrically – see Aligning double sided models). To these is bolted a piece of 3” x ½” aluminium which is skimmed across the entire surface to produce a consistent Z height. Attempting to clamp this in the vices tended to lift it, hence the direct bolting. Come to think of it, direct T-bolting into the table with spacers between would have a similar effect, but might be more difficult to align along the X-axis.

Two stops are bolted to the back edge of this to give consistent parallel positioning for any material clamped to it – the stops can be unbolted if they would interfere with a milling operation. There is a half inch gap between the bottom of the aluminium and the table which enables me to clamp material down using simple strips of ¼” material above and below, bolted together.

     This has several advantages over clamping direct to the table:-

            If I accidentally drive the cutter too deep, it is my bit of aluminium, and not the table, that gets damaged

            I can mount material wider than the vices would accept and still position intermediate clamps (vital for thin material – e.g. 3/16” or less, or long pieces) – attempting to use intermediate clamps direct to the table would limit material width to less than the maximum T-slot gap.

Holding single sided parts                  Author – Derek Mackenzie

            A holding technique I use for parts that can be cut single sided is to leave 5 or 10 thou at the bottom when cutting so all the parts are held together by the thin layer of remaining material. If you do this it is a good idea to glue another piece of material below the material you are cutting – essential in the case of 5 thou as the material tends to “bubble” when it is cut this thin. For 10 thou I glue on a piece of MDF (medium density fibreboard – fairly stable with predictable thickness) using white wood glue (PVA) – this can be soaked off when complete. For 5 thou it is better to use superglue (cyanoacrylate) as you will need a stronger bond – the parts can be removed by cooking to soften the superglue.

            Incidentally if you are cutting single sided in brass you will probably need to skim at least 5 thou off the bottom face first – the rolling process (or whatever it is they use to produce brass strips) sets up stresses in the skin if one side is cut away but not the other, the unbalanced stresses will cause the finished object to roll up slightly. Aluminium does not appear to suffer from this problem (not here in the UK anyway).

Holding double sided parts                  Author – Derek Mackenzie

            When cutting very small parts, I assemble them into a group measuring about 2” square (depending upon the width of material available), cut the first side, thoroughly degrease the whole thing and then fill the spaces with car body filler (polyester resin) with another piece of material clamped on top. The latter has two effects – it provides a smooth level base for the piece when turned over, and it glues the filler firmly to the material – the filler shrinks as it hardens otherwise. The second side is then cut. The complete piece is then put in the oven at 150 degrees C (302 F) for 20 minutes – take it out and quickly peel the parts out of the resin – it is soft at this temperature and will have lost all (well, most) adhesion, but will soon harden again as it cools.
Try not to overcook the filler - it produces an awful pong, makes it difficult to remove and leaves a burnt deposit on the metal

            Be careful when mixing the resin to get the ratio of resin to catalyst right – if too little catalyst is used it will still be a bit soft with poor adhesion and the milling cutter may drag the part out of the resin.

Aligning double sided material for cutting      Author - Derek Mackenzie

Some of the forum messages referred to setting up a jig so that the material can be turned over - this will only work if 0,0,0 is dead centre of the object to be cut. MillWizard has 0, 0, 0 as the near left hand corner (Z0 can be at the top or the bottom of the material).

My program that alters the feed rates prints out the maximum and minimum X, Y, and Z values used, so at the end of the initial roughing pass I move the cutter to X0, Ymax and then cut a hole through the material. You can also determine Ymax by watching the controller display at the end of the pass (for raster cutting only).

The remaining milling operations for the first side are done.

I turn the material over, rotating around the X-axis - see pages 1 & 2 about holding material.

I fit a probe (see note 1 below) into the milling collet, then move the X and Y axes until the probe is just above the hole. The probe is lowered into the hole so that is just missing the material.

A multimeter set to maximum resistance sensitivity (2000K ohms) is attached to the material to be cut and to the probe (see note 2 about insulation and resistance sensitivity)

The CNC controller software is reset to X0, Y0.

The probe is jogged along the X-axis (1 thou inch at a time) until the meter registers contact. The X-position is noted.

The above operation is repeated in the opposite direction. The average X position is calculated, the probe moved to this position and X position reset in the controller software to zero.

The above operations are repeated for the Y-axis.

The X-axis and Y-axis checks are repeated just to make sure – remember the old adage “measure twice, cut once”.

Whilst moving the probe, the miller spindle is slowly rotated by hand to eliminate any eccentricity.

With the X0, Y0 position now fixed, I fit the next milling cutter and move the axes to the expected extremes, and check that Z0 is the same height at each corner – again using the meter attached to the cutter and the material. The material may require packing at one or more corners to achieve this – make sure the material is still securely clamped at one end while packing, or you may lose the X0, Z0 alignment.

Obviously the Z position cannot be checked at X0, Y0 so I usually move the cutter clear of the hole (e.g. to X0.15, Y0 for a 1/8” hole) to do the first Z check.

Notes:-

1. Probe –

I use an old TC milling cutter – I found that a steel probe was too easily bent if I jogged it too far (I occasionally forget to set the jog value to 1 thou, and the default is too large). A 1mm cutter with a 3mm shank has a nicely tapered transition between the two, and I have several old 1mm cutters that broke due to feeding too fast). If your milling collet is not insulated from your milling table you will need an insulated probe.

2. Insulation and resistance sensitivity –

My mill has an anodised table which effectively insulates it from the rest of the machine. I found that setting the multimeter to 2000 ohms produced erratic and unrepeatable readings; setting it to 2000K ohms solved this problem entirely.