PDA

View Full Version : Scalable / Adaptable Gantry Router



andy_ck87028
09-25-2005, 07:13 AM
My need for a mechanical solution started with the realisation that I could not cut things accurately by hand!

As I initially wanted to create a set of railway carriages, it relied upon repeat accurate cuts which apart from not being able to do seemed extremely time consuming and boring.

After researching the web, I discovered CNC. I have had a long time ambition to start a hobby for retirement and model engineering was my answer. I have no training or practical experience in any of the skills needed for model engineering. Building a CNC machine seemed to be a good starting point. (I'll need lots of help in the electrical bits :))

I realized that a CNC machine would enable the repeat accurate cuts that I needed, however I also realized that I could build a tool that satisfied the above need but that tool could also become very limited unless I designed with future projects in mind. So my imagination took over and I thought of a long list of wants that this machine could achieve.

One thing that seemed obvious was that if I build a small machine, I would get frustrated by its limitations and that as soon as I finished the first I would want a second. I also argued that a small machine does not imply increased accuracy or reduced complexity, just reduced outlay. So if the choice was two machines or one machine, then one large machine made better sense even thought the initial investment is high.

That left me with a whole heap of design criteria such as: size of cutting footprint, thickness of material to be cut, type of material to be cut, etc.

Furthermore, I realized that these machines could wear a whole range of adaptations such as lasers, air brushes for painting, die grinders, belt sanders, orbital sanders, jigsaws, etc.

Also, I could experiment with cutting glass, stone, metals, plastics, etc.

I was not limited to 3 axes either and could incorporate a turntable or a swivelling head, etc.

I opted for a design that would take a standard sheet of material made of hardwood or aluminium whilst keeping my options open for more challenging materials. I also felt that 6in travel for z axis was a reasonable challenge, whilst allowing for additional axes to be retro fitted.

I then reasoned that a standard sheet of 2400 * 1200 seemed a lot of investment for something that was still basically a hobby. It was then that I decided upon modularisation. I decided to build a machine that would take 1200 * 800 with the ability to be converted to a 2400*1200 (that is 1200*800 + 1200*800 + 1200*800) easily.

With a modularisation target in mind, I concluded that I would have to have a gantry design as then, the only significant change from 800 to 2400 was the lengthening of the x rails. With a gantry, I am able to put the x, y and z motors all on the gantry.

So, I was committed to a footprint of 800 (x) * 1200 (y) with a 6inch z travel. That led me to be concerned about structural strength. I thought that a gantry supporting a router assembly spanning a cut of 1200 would require reasonable strength. Hence, I decided upon mild steel for all assembly. Apart from anything else, I wanted to learn how to work with metals including cutting, tapping, welding etc.

Also, I didn't like the idea of aluminium. I didn't think it would be strong enough and I was concerned about prolonged use introducing fatigue. As a person who knows very little I might be wrong on that consideration. I also thought aluminium would prove more expensive. (Still don't know ... haven't compared!!)

So, by deciding upon a machine of this size, I did not find out until much later that I had inadvertently committed myself to at least gecko level controllers and 400oz/in motors.

I had a long look at steppers and they seemed to be understandable. I had a long look at servos and they still seem like an absolute minefield.

Also, I had a long look at drive gear and compared threaded rod with timing belts. Timing belts seem a lot simpler for me so that is the path I have taken. Furthermore, with the possibility that my machine might be expanded on the x axis from its existing 800 to 2400, timing belt extensions are far practical to extend than threaded rod!

Also, it had to be elegant. I wanted to create a machine that looks simple and is a joy to watch. That means that I had to ensure that it looks balanced, designed with simple structures of triangles, circles, lines and rectangles

In summary, I realized that the design criteria I had set was for it to be elegant in design, robust, functionally effective, accurate and modular.

So, I have started this log to show everyone what I have created.

So far, it has consumed a heap of time (both thought and build) but when I compare with other logs, not that much money. Mild steel is cheap. I reckon the cost of the frame and gantry and head and router but excluding motors, controllers, drive gear is around $800 Australian. Three geckos ($450), 3 400oz steppers ($400), control box ($150), timing belt and gears ($200) will keep my investment to a total of around $2000 which I reckon is pretty cheap.

Furthermore, changing it from a 1200*800 to 2400*1200 machine will probably only cost me an extra $800.

Considering a finished model railway carriage in Gauge 1 could cost me $2000 then I reckon I'll be in front very quickly.... plus I'll have had the satisfaction of getting this machine finished and be able to use it for heaps of other projects.

At this point in time, I have been working on this for about 9 months. A lot of it is in semi assembly but I have now started to go back over it all from the beginning with a view to bedding down prior to proper assembly.

I have got some pictures and will soon get the first group on this thread in the next few days.

I hope this log will be of help and look forward to any feedback.

Andy

OCNC
09-25-2005, 08:48 AM
Looking forward to seeing what you've done!

andy_ck87028
09-26-2005, 06:21 AM
Attached are 4 pictures of the frame of my router.

The first picture shows the basic frame. I decided to have it as a low level frame to make it more portable ensuring that the rails and gantry are removable thereby allowing the basic frame to be stored when not in use.

I have been concerned about transmitted vibration and have therefore contemplated putting it on a solid table or alternatively extending the legs of the basic frame. If I take the latter option, I could attach proper feet. I have yet to decide.

As you see, the basic frame is merely a box with four posts on each corner. Those posts will take the X rail supports.

After a while, I realized that the frame had very little torsional strength. Furthermore, angle iron can flex in its own right and is less rigid than box section profile. I decided to strengthen the basic frame by introducing triangulation onto each corner and infilling the shape with particle board which is glued into place. I have taken the bracing one stage further and have glued particle board around the outside. The basic frame now is a lot stronger, has much reduced torsional flex and is still quite light. The strenghening process was quite cheap as all it took was a bit of particle board and 8 pieces of 20*20*3mm angle iron and a few bolts.

You will see in the second picture that I have placed the 'x' rail supports onto the vertical posts.

You may recall that this machine is modular. All I need to do to change it from a 800*1200 to a 2400*1200 is create more rigid posts and increase the 'x' rail (and support) in length from 800 (+ 250 mm overhead) to 2400 ( + 250mm overhead) and extend the accompanying timing belt.

With the basic frame and triangular strengthening, I have left sufficient room to accommodate a turntable.

The working height of the machine will be achieved by placing a series of bars across the top of the frame and bolting each bar onto the top of the basic frame. Those bars will then be about 250mm below the top of the vertical posts which will alllow enough room for the gantry and router assemble and still enable 6in z axis travel.

The vertical posts are 40*40*4 square section. The frame is a mixture of 50*25*2mm rectangular section and 40*40*3 angle iron.

All pieces have been bolted together with enlarged holes to allow adjustment. Welding will happen (maybe!! - I'll see if I need to first). Instead of welding, I could always use epoxy glue and bolts.

Andy

OCNC
09-26-2005, 09:14 AM
I like the modular concept but with all of the bolts and oversized holes I'm wondering if you aren't going to have problems with rigidity.

andy_ck87028
09-26-2005, 05:48 PM
I'm open to more feedback on that as it is something that concerns me also.

I have a number of options open to me.

Do nothing
Weld each joint once positioned and bolted down (would a tack weld be sufficient?)
Unbolt every joint; then bolt up with epoxy on each surface
Bolt the frame down onto a rigid floor and have the floor impose the rigidity factor
30 coats of paint (lol)!
Incrementally solve the problem


Taking the incremental approach further..........

The introduced chipboard will definitely improve rigidity on a number of planes however given that I am going to put a working surface over the frame by placing multiple steel rods across (say 5 lengths of square section 30*30*3 spaced evenly) then I might find that by only welding those down it could be enough to give me the rigidity I need. I could also weld corner plates on and that might be enough.

All opinions welcome.

Andy

OCNC
09-26-2005, 07:56 PM
I'm open to more feedback on that as it is something that concerns me also.
Andy My thought would be to continue with the design using the nut's and bolt's type assembly and then look back and see where the different elements could be glued or welded to gain rigidity while maintaining maximum access to modularity for future expansion. Using epoxy vs. welding would ease revisions and may well be strong enough when combined with the bolts. Use tack welds for greater durability where you know they won't effect the modularity. As you go from a small machine to a big machine the forces will increase directly form more powerful motors and spindle and indirectly from longer lever arms. This may require thinking in terms of a maximum case condition early on which may be difficult to do on a first machine.

andy_ck87028
11-12-2005, 02:48 AM
Hi Guys

Exciting day for me today. I spent the day cutting and bolting my y rails.

X rails are assembled as is gantry frame, gantry trucks and head assembly trucks and head assembly itself.

So, I now have a head assembly that moves in two planes. :rainfro:

The router assembly is positioned in place and now I have to put in the z rails. (4 already purchased using silver steel rods and will be placed on each corner of the router with movement achieved through bronze bushes (teflon coated))

Am aware of lots of concerns regarding my design but maybe not all ..... so feel free to comment.

Happy CNCing

Andy

Rance
11-12-2005, 09:35 AM
Andy,

Thanks for the update. I'd like to see how you are supporting the X axis rails. Do you support them just at the ends or the entire length? Please keep us updated as you make progress. It inspires others like me and gives us new ideas.

Rance

andy_ck87028
11-12-2005, 04:47 PM
Rance

Hopefully these pictures will give you an idea.

If x sags, I can always put a torsion box underneath. The beam is 40*40*4mm

Regards

Andy

Rance
11-12-2005, 09:21 PM
I like the way you tied the rod down, looks very sturdy.

mikie
11-17-2005, 12:48 AM
Hey Andy,

I now understand what you were saying this morning with bolting the Stainless to the two plates. I don't think you'll have any issue with the angle iron flexing and causing you problems.

So the pipe is only secured at each end?

Cheers
Michael

andy_ck87028
11-17-2005, 04:13 AM
Mike

Thanks for your encouragement.

The stainless is supported from underneath along its full length with the angle iron, square section and plate sub-assembly shown. The stainless is pulled down onto that sub-assembly repeatedly along its length(on average one stainless bolt every 200mm).

Both x axis and y axis rails are constructed the same way.

Any stainless tube / mild steel expansion differential will be accommodated without weakening the assembly. Welding / glueing can occur throughout the sub-assembly with the exception of where the stainless tube is bolted down onto the mild steel sub-assembly.

Looking forward to you and Nigel having a looksee.

Andy

ynneb
11-17-2005, 06:12 AM
Andy, I must say I am smiling at what you have done.
Firstly, I remember you telling me how you were going to build your machine out of Mecano ( I said i liked a guy who was prepared to try new things, while thinking I didnt hold much hope for it) Then you were going to make it out of an old Triton work bench, I was thinking, yeh I am glad this guy has dropped the mechano idea. At least he has a better chance with the triton, but still not the best idea.

Then we discussed welding and glueing of you metal frame. I hated the glue idea.
I see you have settled for bolts. I have previously offered for you to take my welder, the offer is still open.

Its fun to see your progress, because I can see how you have considered many options and you project has evolved greatly over time. When you are finished, I am sure you will look back with fondness over the time you spent thinking and considering options for you machine. Its a head ache in the mean time though, but your progress is going well.

Keep up the good work.

andy_ck87028
11-17-2005, 03:58 PM
Benny

Thanks for your very kind remarks.

And thanks for repeating your offer over the welder.

Much appreciated and I will most probably be taking it up.

I'll hopefully pop over soon and who knows Santa might have a beer with him :)

Andy

Pat2000
11-17-2005, 04:48 PM
Hey just catching up with yer thread Andy... interesting... looks good! I'm interested in that Z axis when you get there, thanks for the PM enjoy the heat you lucky Auzzies.

Pat

andy_ck87028
11-20-2005, 02:42 AM
I've thought of two innovations this afternoon that I will be incorporating into my machine.

Thought I'd report on them as they might be of interest to others.

Large Z Travel

From the pictures below, even though the pictures aren't that clear, the router is a long way down from the y rail. The ideal place for the router cutter has to be as close to the x and y plane as possible as the further the distance away from those planes , a) the greater the risk of vibration, b) inaccuracies caused by increased moment c) greater wear on parts and d) accelaration/deceleration problems caused by the gyro affect of the router.

The problem is that if you want large travel on the z axis, the greater that distance can become. Furthermore, if the head assembly is designed for a 6" Z, then the normal place for cutting will be at the bottom of the travel which is the worst place from a design efficiency point of view.

So, my answer to this problem is 1) keep the x plane as close to the y plane as possible, 2) have the normal cutting position for the z immediately below the x y plane 3) still allow for 6" z travel and 4) have a variable height cutting table. When the 6" Z is called for....... lower the cutting table.

Modular Head assembly

From the pictures you will see four vertical angle irons within the triangular frame. Try to visualise that as two components, 1) a self contained box (called Z assembly) and 2) a triangular head frame. The box then goes within the head frame.

Earlier today, I was thinking about structural rigidity of that entire assembly and realised that with the Z assembly being a self contained box, then a) I would ensure that assembly is rigid, b) alignment is then managed purely as a positioning issue of the box assembly within the head frame (assuming the box itself is square) and c) I can easily remove the Z assembly if I need to replace the tool attachment.

To achieve that, I would need to ensure the ball screw is attached to the head frame as distinct from the Z assembly.

Effectively what I am describing is the separation of each sub assembly into discrete parts and then ensuring that the purpose of each sub assembly is clearly defined.

Hope those thoughts are of interest.

Cheers

Andy

andy_ck87028
06-26-2006, 08:11 AM
No other pictures at this stage but plenty of effort.

Many false starts but still having fun.

Have bought 3 of these Nema 34 637oz steppers and they are coming back from O/S in my wife's suitcase (I'm not popular!)

I've just bought my z axis bearings , have got my Z axis shafts (from a scanner), have just bought 5 metres of timing belt and a couple of pulleys.

Have bought a toroid.

New phase of building has just started.

X linear rails built, y linear rails built, trucks assembled.

Having fun :)

Hope to be finished by Xmas !!!!!

Andy

NB Thanks for the patience almost everyone has shown in coping with my often inane questions. ... just wait when I can't get the b----y thing working :)

andy_ck87028
02-24-2008, 03:51 AM
Well 18 months later and the machine is now manufactured.

I am now going through the process of aligning and tightening; bedding down, etc.

I have also just got me a camera :)

So, I thought that I would show some pictures that will reveal the detail of my design.

The two pictures attached are one of the carriages that sit on the X axis. Attached to the carriage is the sub assembly that the y rails are bolted to.

More pictures soon.

andy_ck87028
02-24-2008, 04:30 AM
Oh yes.......

Some of the earlier pictures of the head assembly are now history. Bin fodder!!

I'll get some pictures up of the finished version soonish

Andy