[ImanCarrot]

I would really like to know how to sharpen these tools. At the moment I have to send them away for relap which obviously costs, but the delay in getting them back is the killer because I obvously have to X-Centre, Height Set and set Radius again, all of which can take a day (I'm machining optical lenses).

It would be neat if I could lap the tool in situ when it becomes worn.

Any ideas greatly appreciated.

Many thanks in advance.

[NEATman]

Imancarrot-
Could you post a picture of the tool? I have designed a built a diamond tool lap, and air bearing diamond turning lathes, but the shape of the tool, and the machine you are using it on make a world of difference.

I would guess that you are running a Pneumo Precision, Precitech, or simmilar type ultra-high precision lathe. I know that Precitech offers a cutter-comp add on package that compensates for the radius of the tool tip, and modifies the toolpath to get a more accurate cut.

Maybe what you need is a "kinematic" toolholder. One that repeats to extremely tight tolerances when you remove and replace the tool. This same mounting pattern could be used when you mount the tool for lapping. There would be slight material loss during the lapping, but if you had a electronic LVDT indicator you could measure how much material was removed by the lapping. What radius are you trying to create on the tip of the tool, and how accurate does the radius need to be? Do you have any idea what process is being used now to lap the raduis on the tools?

Neatman

[mc_n_g]

I would be too scared to relap the tools myself due to possible bad finish. I am assuming you are talking about 'real' diamond or CBN rather than diamond shaped tools. I have no idea of the surface finish required but I am sure it is very high. Another assumption (bad on my part) is you are CNCing these lenses. You have potential to 'throw off' your cutting edge by rounding it, removing the true radius, etc.
I have no idea as to what level of optical lenses you are dealing with but any abnormality in the lense can cause sperical abberation which will cause the lense to fuzz up due to the rays not converging. It really depends upon the specs for the lenses. If you are trying to save a few bucks because it is eating your profits or it is adding to the timeline I have no idea what to tell you. Perhaps renegotiate the contract?
Tool grind/lapping machines are not cheap (there is a reason - tolerances). You might try to find a used optical tool grinder with lapping capability. Another option might be a surface grinder with a speicalized jig and costly specialized lapping wheel.
Either way you need to be able to control the lapping process.
Good luck,
mc_n_g

[ImanCarrot]

thaks NeatMan and MC G N, for your reply :P

You are entirely correct, it's a CNC ultra thigngy, airbearing spindle and vibration isolated etc. It's an Optoform, although I've managed to get it behaving properly like a Precitech or MSG (after a few choice words lol)



I already use tool radius compensation and I can correct for the radius given by the LVDT probe no problem. It's wear when the tool begins to give a bad finish that snookers me (typicaly caused by slight chipping at the diamond cutting edge).

These tools are single point natural diamond (not synthetic) and are like... well, think of a stack of coins pushed at a slight angle (a skewed cylinder) with the top forward "coin" being the cuting edge. So what they do when they re-lap is to remove this top "coin" and you have a new cutting edge with the same radius.

Now if I could do that process in situ (obviously raising the tool by the amout removed too) it would save me a lot of time.

It's not the profit margin I'm worried about, but the time taken in setting up the tool which can take anything from an hour to a day and is frustrating

[NEATman]

ImanCarrot-
I am farmilliar with that type of tool, and the cutter compensation LVDT. I think that it would be quite difficult and messy to lap the tool directly on the machine. I think that your best bet would be to design a tool holder that is removable, but very repeatable. The best way to describe a kinematic mount like this, is just like the mount for the LVDT probe. Six ceramic rods, and three ceramic balls, and a way to clamp it in the center repeatibly with the same force. This same mount could be incorporated into a lap fixture. You would then also need vertical movement to accomodate for the new tool height.

Before I go too far with a new tolholder, which adjustment is causing you the most difficulty? And how accurately do you need it to repeat? If the height and radius are not the concerns, then the XZ position must be the problem. You could make an additional toolsetting fixture that mounts to the spindle on the kinematic mount (for the LVDT) and measures the location of the tool. Perhaps two more LVDT's, one for X and one for Z, so you could simply mount the tool, adjust the height, and bring it over to the two LVDT's, and measure the offset. That new offset should be easily programmable into your tool settings, and that should get the tool within about .00005" if I remember how well that kinematic mount repeats.

The more I think about it, that LVDT tool compensation system has the probe mounted horizontally, so you already have the Z direction done. You could set the Z location of the tool with the LVDT you already have. If you were to get another LVDT, and make another mount like the one you have, except that the probe is mounted perpendicular to the current orientation, you would have an X offset value. Athough due to the geometry of the tool, 90° may not be the optimum angle, as it would miss the radius on the tool. You could mount it at an angle, so that the probe tip still contacts the radius, and do a little trig to figure out the X offset. If the height is also a problem, you could perhaps make one more fixture with another electronic gage head to measure the height of the tool, and dial it in. If I remember correctly, the toolholder you most likely have has a differential thread and knurled knob at the rear, and will adjust the height very accurately.

Neatman

[ImanCarrot]

Neat answer Neatman you certainly know your stuff!

The LVDT probes I have in X and Z get me roughly set for X and height (to within about 5 microns or so). I can then adjust height further by checking the central pip under X50 magnification and moving the tool in Z accordingly. Adjusting X is a bit trickier but after measuring the form of a test sphere the error (either M or W shaped) tells me which way to jog X. I then do some big sums to find out how far to jog and ignore the results and just guess lol- this iterative process is just as good

However... it can take all day to get two tools setup then you find out that one is giving a bad finish and needs relapping.

Thinking about it... even if I did find out something to lap diamond (thinking it would need to not contain Iron or carbon to prevent graphitisation of the diamond), there's no way to do it by hand and keep the top rake controled which would be critical to maintain the form of the tool as a perfect sphere- I'd end up with an elipse or something which would directly replicate to my parts.

ahh well... was worth a go I spose. Thanks again for taking the time to reply

[NEATman]

ImanCarrot-
Well there is a good reason I know that stuff... But I'd rather not say.

Anyway, I guess that you could still work on a new design for the toolholder, but it would have to repeat to about a micron. Not easily done, but possible. Also, you would have to work with your vendor who does the lapping to create a fixture that uses the same socket to mount the tool at the proper angle.

One other variation that would use your standard tooling. If you created this highly repeatable toolholder, you could have a mating "socket" on a bench, so that you could align the tip of the tool in all three directions using a 50x or 100x microscope, and electronic gageheads. As long as the toolholder repeats well, this should get you within about a micron of position. I would still reccomend a test cut after you install the tool, but you could set up multiple tools and maybe cut the iterative process down to one ot two, instead of a full day. And it would not require special tooling at your lapping vendor.

If you are interested in this approach, let me know, I can probbably help with a design.

Neatman

[ImanCarrot]

Tanks Neatman, but the lapping will snooker me I think- I'd need to have some ceramic or synthetic ruby lap angled at the tool's top rake (typically 5 degrees) reciprocating back and forwards with say Engis or Hyprez Diamond Compound. The mess and setup entailed would probably be just the same amount of hassle as setting up a totaly new tool heh, but it was nice chatting about it. Like I say, you know your stuff!