Making a carving copying machine

The need for a copying machine stemmed from a desire to get away from the typical ‘board cut’ style slingshot which does not fit the hand.  So as a part of my  theoretical Island project, I could see the need to offer a more ergonomic form which cannot be simply produced with just a saw and a router. Another problem to be addressed was that one size handle does not fit all people. A woman's hand is generally smaller than a mans and a child’s smaller still, so it seemed logical to address the question of grip size.
Another factor is that some people are tight handed and others left and an ergonomic handle designed for a right handed person, will not fit a left handed person. So we are not just looking for a better shape, but they also must come in left or right hand and be available in several sizes.  The question was how to do it quickly and cheaply, to make it economically viable. I think (at the moment) the answer is to use a copying machine.

Copying machines in general.

You can find loads of video’s on youtube showing copying machines, that carve a replica of a ‘master’ carving. A metal ‘stylus’ follows the detail of the master, repeating the same movement at the cutting head. Generally the difference  between manually operated machines is the number of axis they work in.  A two axis machine  can move the cutting head  sideways and up and down. A two axis machine works OK if you a reproducing ‘flat’ masters, such as a plaque. For replicating three dimensional object such as an ergonomic slingshot handle, you generally need a three axis machine  that can also rotate around the master.
Now it just so happens that making a three axis machine isn’t much more difficult than a two axis machine and if needed a three axis machine can  easily be restricted to a two axis machine. So the plan (at the moment)is to do just that !.

Construction techniques.

In the past I usually have  metal tubes and plates, secured together with pop-rivets, which is the technique used to construct aircraft.  When the slingshot project appeared, I could see that a number of jigs and simple machines might be required to do what I had in mind and for a change I wanted to weld them together.  Unfortunately, although I sent myself on a MIG welding course about five years back, I have not had any practice since and do not have the time to relearn the required skills.  So to fix this the current construction project is a MIG welding jig, designed to eliminate the need for hard earned skill.  That will be finished this week and then I will use that to construct a copying machine (carving machine) using  20.0mm square steel tube (as it turned out … I didn’t) .

Machine size.

My machine only has one task in life and that is to make slingshot handles, which are not large, so the ‘machine can be quite small, depending on how many carving spindles it has.  I think the sensible thing to do is keep it simple and have only one spindle for the master and one for the carving blank.

Carving speed.

How long it takes to carve and sand a handle will depend on several factors.  One is the compromise between how much of the finish will be produced in the machine carving and what will be produced in the sanding phase.  Basically the quality of machine carving is proportional to the amount of time covering every part of the wood with the cutter.  The time taken to do the sanding depends upon whether we do it under power or manually.  So in the next few days I will be juggling with this compromise in my mind, because the answer may determine the actual form required of the machine. The point is, I only want to attempt this machine once !.

15th September 2012.

I have done some initial thinking and basically a  tool / machine breaks down into two parts.

The stylus / cutter assembly (two, three or four axis).
The master  and blank rotation assembly (one axis).

The main problem I see is the shape of the finished carving … it makes things awkward as there is no easy way to secure the blank onto the work spindle. The same problem means that we cannot put the master and the blank onto the same shaft. So in effect we are looking at two inline shafts, one for the master and one for the blank, connected together with gears, toothed pulleys etc so that when the master moves, the blank moves with it. Since I am right handed I would want the master on the right of the machine and could use the left hand to rotate the master / blank on their axis.
The above involves at least four gears / toothed pulleys and a  hand full of bearings and I was beginning to wonder if it was going to be  an economic proposition (since the island project is theoretical (unfortunately).  However I had a rummage in my workshop and found the items below …

The pulleys are 40.0mm in diameter and 15.0 mm wide. I did some thinking and for a dozen reasons, decided on a layout for the master / blank rotation half of the machine.

Looks complicated, but I could not think of a simpler way considering all of the problems involved. Rotating the left hand knob rotates both shafts in the same sense.  You will notice in the top photo there are five  toothed pulleys !.  Well I had been wondering if the actual turning of the shafts could be done under power. To do that the fifth pulley would be fitted onto the lay shaft between the other two pulleys and go back to a very low RPM motor. At first glance it might be thought that it would not be possible to do this and end up with identical handles, but I have worked out a way for the shafts to dwell at any point in order to get the cutter down to the correct depth.
In practice  the master and the shaft cannot be mounted onto a straight shaft (as shown), because of their shape, but I think I have a solution for this too.  The master has to be easily changed because of the need to make different style and size grips.
In the thinking stage I also have to work out what benefits would be gained over manually sanding the grips to shape, sometimes the simple way is the best.

17th September 2012.

Following on from the sketch, I decided to make the two brackets that will hold the master and blank shafts (separately).  These shafts are  going to be turned relatively slowly, initially by hand. Also the cutter will not be imposing much of a load either, so the two brackets need not be terribly thick. In my case I was limited to the steel thickness that I can actually bend.  The hard ware store only had one thickness in 25.0mm steel strip, so there wasn’t much choice in the matter.  I had in mind fitting an internal brace inside each bracket, but do not think it will be required.

In order to get all the holes lined up I decided to bend the metal first and then drill the bearing holes (8.0mm for an 8.0mm shaft).  They came out quite nicely and the photo shows them both on one length of 8.0mm round steel, so that I can check alignment. Actually alignment and play are not really issues on a copying machine, so long as any movement affects both the blank and master together. To make the assembly as compact and user friendly as possible, the master shaft will now be on the left hand side. That now gives plenty of space for mounting the blanks and removing the completed grips.
Initially, I had envisaged  the complete copier to be quite small, until I took into consideration the need to be able to shape  board cut forks with a three layer handle !.  So I will take that into account and move the above two brackets a bit further apart.
The next job will be to make up two brackets for the lay shaft. I may need to turn up a knob for the left hand end of the master shaft for rotating both master and blank together. That will obviously be manipulated by my left hand and my right hand will be guiding the stylus on the master (I am right handed).

18th September 2012

Made up the lay shaft support bracket …  in one piece to make belt tensioning easier, whilst maintaining alignment.  
At this stage it is just a jumble of pieces and I needed to know if when assembled, whether the whole assembly was going to be rigid enough for me.  So I have temporally mounted the bits on a scrap piece of board, to find out.

The left hand short spindle is for the blank and the right hand one for the master, which will be mounted on the extreme right hand end of the shaft.  Things that are missing at this stage is the left hand knob to turn the shafts, collets to position the toothed pulleys and a means of mounting both master and blank to the 8.0mm steel shafts. I have mounted the lay shaft bracket lower than the front two, to try and clear as much space as possible for the cutter arms assembly.  Effectively this assembly add’s a single axis to the complete machine, the other axis of freedom will have to be provided by the cutter and stylus arm assembly.
I think that I have left sufficient space between the front two bracket to get a complete sling shot master in  to shape  54.0mm thick three layer multiplex board blanks.
Using the modular approach, I have left myself plenty of options.  I think the next job is to think up some way of holding the masters in place on their  shaft, which relates directly to how the blank will be secured to it’s shaft.

19th September 2012.

Securing the master to the shaft.  Actually, any solution for the master was also valid for the blank holder. I woke up with an idea of how to do it, but wasn’t sure if I could pull it off. The first problems was I needed a short piece of 25 x  25 mm BMS angle iron and the hardware shop had none. In the end I found a very rusty chunk of 50 x 50 mm and decided I had no choice but to cut what I wanted out of that. I wasn’t looking forward to it because the metal had deep pitting, I had to get rid of somehow.  Anyway a few hours with a hacksaw and file produced the following ….

I imagine that I will make the blank holder in exactly the same way as I am quite pleased with the end result.  It has been designed to accept 18.0mm  thick master, or either the board cut handle of the multi layer ergonomic hands with forks. All parts were made from mild steel because I wanted to silver solder two parts together
The shaft is the 8.0mm shaft I made yesterday and I turned one end down to 6.0 mm. The reason for this was that I wanted to rivet  the shaft to the angle component and then silver solder it as well !.  That all worked out rather well and it is a very strong assembly. In practice it acts like a rotating vice to hold the master, or blank.  It does mean I now need to make up some new templates  with a  small square extension on the top end that will actually be secured into the ‘vice’ and saw off later after shaping.  Any master will also have this square extension.  
Why did I do it this way ?.  Well to do it otherwise meant I would have to be able to slide the master shaft out  to change masters. This way it can be done with the shaft ‘in situ’ and that means the machine can be a bit narrower.  
Angle iron, even bright mild steel very seldom has a 90 degree angle on it and so it was with the junk piece I used. That meant that after I had fixed it together, I then  had to adjust the angle slightly (with a big hammer !).
Next problem is to find a large knob about three inches in diameter. I wonder if my home made wine is ‘drinkable’ yet ?

20th September 2012

Woke up this morning morning realising I had made a mistake.  It concerns the way we use copier carvers. In my design one hand turns the stylus  / cutter shaft using the knob, and that is my left hand sorted out. The right hand is on the stylus, but we don’t actually have to watch it, because in practice we can feel what the stylus is doing and also the master.  What we do have to watch is the cutter on the blank, to make sure we have moved the stylus / rotated the blank … So that there are no spots that the cutter misses.  It would therefore be logical to put the blank spindle in the centre of the machine and the master spindle  on the right hand side of the machine. OK, I think I finally got that sorted !.
After making the shaft yesterday I realised that it was plenty strong enough for me to simplify, or in other words, I had forgotten the need to profile three layer board cut handles with forks !. I had to make a second shaft anyway today, so I changed the design a bit to accommodate a complete handle master. I like the finished item so much I think I will modify the other shaft to be the same.  Basically the change is from ‘clamp on’ to ‘bolt on’.  It solves some other things too like registration of the blank, changing masters, variations in board thickness etc.

It means that the blank will now need a 6.0mm hole drilled in it. To make sure this is accurately placed, it will be included on the marking out template. That automatically sorts out registration for both master and blank.  
22nd September 2012.

Made up the small grip template, that includes the ‘sacrificial square’ for clamping purposes.  

Just won a three inch aluminium Picador control wheel, for this rig on EBAY …

This is for the left hand master shaft …. Which,  via the lay shaft, also drives the stylus shaft. It has a half inch hole so I will have to make a bush to step down to 8.0mm.
One problem with maintaining options is that at some point in time you have to make a decision, on which one  you are going to employ.  At the moment I have four possible methods of making the handles and will not know which is the best (ie fastest, cheapest  and easiest, with the best end result), until I get some multiplex board to try them all out.  It may mean making up two different cutter arm assemblies to try them out.
One problem I am looking at, is that the tip of a router cove bit is not the best part of a bit to use and where-ever possible, the major diameter of the cutter should be used. The reason is that the tip of a round ended cutter scrapes rather than cuts and the nearer we get to the major diameter the faster the actual cutting rate.  So cutting at a 45 degree angle is better than a simple vertical plunge cut for carving. This tends to complicate a three axis design.
Another factor is the wood removal rate against the degree of finish we require, or in other words just exactly how much hand finishing do we really want to do ?.  

23rd September 2012.

Decided I need to start on the cutter frame assembly, but have no idea as to dimensions at this time !.  So to get things started I made up one vertical element this morning … which I can shorten later …

The tube is steel 20 x 20 mm, the brackets are steel too.  I will make the other one tomorrow.  Since the height required is an unknown, what I may do is make the height adjustable as I did with the MIG welding jig.
For the  three axis version,  the frame that holds the router or cutter, will pivot between the two uprights, which gives me the second axis and the router /cutter will be able to slide laterally to give the third axis. The first axis is provided by the  master / stylus  rotation  assembly.
I could mount the router on the ‘inner’ frame but then I would have to add a hefty counter-balance weight at the rear of the frame to  counter the weight of the router. Another possibility is to use a flexible drive powered by an ‘off frame ’ motor, but I am unconvinced that the flexible drive  I have just bought would be ‘man enough’ for the task, even though it has a 6,000 RPM spec.  The handle part  would be bolted to the frame. The handle really needs two ball races in it.  Another possibility is to mount the motor / router on the  pivot line of the frame ( Frame  C of G) and use an extension shaft to connect to a  ball race mounted cutter, on the front of the frame.  We shall see !.

24th September 2012.

Made up the second upright and drilled the adjustment holes.  The gods were out to get me this morning and half the holes wandered off, so I wasted an hour moving them back again to where they should have been in the first place (if you don’t know how to move holes you are in for a hard life !).

Also decided on some sizes for the inner frame and made the side members 350.0mm long, with a 250.0mm swing. Actually by the time I get the first lateral frame members fitted and allow for the length of the cutter,  the swing will be about  300.0 mm. This is important because the arm (and cutter) describe an arc, which means the centre point of the cutter effectively moves backwards and forwards depending on the angle of the cut and the trick is to make sure the major diameter of the cutter always stays in contact with the blank as it is profiled.
Out of all of the options for the actual cutting part of this assembly, mounting the router under the arm pivot line and making up an extension shaft for the router bit, is starting to look the best option … at the moment. This would put the router shaft at 90 degree’s to the classical vertical plunge angle, but that may turn out to be an advantage for this application, as the finished shape of the  handle is somewhat challenging to machine (well for me anyway Hi!). The run - out of the extension shaft is going to be critical considering the very high speeds involved.

26th September 2012

I now have a single lateral frame member mounted on the two 20.0mm tubes that are shown in the last photo and that allowed me to  work out how wide the traverse  assembly needs to be  (600.0mm).  For the traverse I have used 1/2” steel rod, which I just happened to have and it will be more than strong enough for the job. To allow the frame to slide sideways freely I made up two Delrin roller bearings which look something like pulley wheels and they will roll along the top of the 1/2” rod.  Turning the 1/4” radius grooves turned out to be a time consuming task, that I was glad to put behind me !.  I also made up four angle brackets that will be pop riveted to the frame pivot points to fix the Delrin roller bearings in place.  Hopefully I will complete that tomorrow and then take some more photo’s.
Once I get the arm functioning, I can think about how to mount the router and cutter extension shaft.  One problem is that the traverse part is competing with the same space as the router, so I am not 100% sure which way I will go at the moment.  The process of adding the router to the beast, will introduce at least  one more lateral member to the frame.  I may have to add a third,  to support the weight of the heavy router.  I will just have to ‘suck it and see’ where the C of G of the horizontally mounted router actually is and then fix the bearing rollers on that line.

27th September 2012.

I usually go into the workshop with an idea in mind … switch off and hours later something has appeared on the bench.  This morning I had a eureka moment and suddenly realised that  one does not need a copying machine to make the handles I have in mind. It is often like going to the wizard of Oz, asking for a copying machine and he gives you something completely different, but it all makes sense.  Instead of a multi axis copying machine, I simply need a special template and a slightly different approach to getting the initial shape.    Two attempts later, I have my template !.


I have to make up a ‘pricker’ centre punch for it, but I think it will work OK. The only question now is what do I do with all the parts I made for the copier ?.   The Picador handle also arrived today, so I turned up an adaptor (½” to 8.0mm) bush  for it.  It was listed as being a 3” handle but turned out to be a four inch diameter one . If I use the copier components I may have to use a spacer block under this to raise the handle to a useable height, another inch ought to do it.