Thread: Hi from Guildford
Just saying hi from the sunny climes of Guildford.
I've built one machine without any real background or research into it (I did it as a bit of "let's see if it works") - and as a result, didn't do it right :) (It worked, though - just not as well as I wanted)
Now doing some research for the next version...
Hi tribbles, welcome to MYCNCUK. Please take the time to view the FAQ section for answers to questions you may have, if your building a new machine be sure to start a new build log detailing your progress !
Thanks for the welcome.
I do have a site for the old machine, but it's too embarrassing to publish :) (It was way to big [there was a reason for that], and as a result, way too unstable).
I'll certainly be doing one for the new machine.
05-04-2009 #4Nothing is foolproof......to a sufficiently talented fool!
Well, one of my hobbies is building cars - about 18 months ago, I finished my latest one, and I've had a hankering to make one myself. So, I designed a new body for the existing chassis (this, for those of you who don't know the regulations means it doesn't need to be re-tested as a new car). I approached the manufacturer of the kit with the designs, and he was (and still is) interested in making it.
However, he wasn't convinced that we could take my design and get it carved out of foam (ready for the mould to be taken) - certainly not without £20K+ spent on doing it.
I persuaded him to have a go, so I set about designing a machine that could do it.
The machine would need to be a 5-axis CNC machine in order to get in most of the nooks and crannies, although two of the axes could be manual.
The initial design was for a 5x2x2m machine, with the cutting head suspended - I agreed with him that I'd build it in his workshop (I don't have such a space available to me); I'd fund the electronics and such, and he'd fund the mechanics (it didn't end up that way).
When we started, it was scaled down to be a 2.5x2x1.8m machine, since getting a 5m length of some kind of threaded bar was too expensive - we'd do a third of the car at a time.
About a week and a half into the build, we decided that we'd cut blocks of foam 1200x600x100mm and stack them to form the body, but as we had it mostly built, we'd use the existing frame. This also meant a 2.5 axis machine was required (some of the more intricate bits would need to be done by hand).
Unfortunately, while accuracy was good enough (the nearest mm is fine; I was getting conceptually better than 0.01mm), speed wasn't (it would take something like 700 hours to do a whole car). Increasing the speed was possible (and I did spend some time doing it), but then the size of it was causing problems with vibrations, and accuracy went out the window.
So I did a quick bit of research (well, looking at YouTube videos :) ) and found a design that was ideal for the size we're now doing. So, now we're happy in what we need to achive, I'm more-or-less ready to start.
In summary, the broad design was right for what we had originally aimed to do, but we chose inadequate steel sizes - and when we went to the smaller size cutting, it was just too big (there was also a requirement for it to be movable - although I'm trying to get him to relax that :) ).
I have to say that it was a very good learning exercise - I'd not done anything like that before (except for a "scanner" I made at school some 20 years ago - that had a resolution of a Lego brick).
ETA: I have seen yours, Smiler, and the new design is fairly similar to that, although the bed was going to be closer to the X rails. It might change :)
The first one is never right :)
Since I started mine, I have replaced the unsupported rails on the X with 25mm HIWIN rails and the entire Z axis is now steel. I was getting slight tramlines in really dense materials which I could only put down to the rails and the flex in the PVC. I bolstered the PVC with Ally angle down the sides but it didn't really reduce the flex much so I had to go nuclear and replace the lot with steel. Solid as a rock now :)
I have approx 200mm of height on the Zaxis simply because I want to add a fourth axis sometime soon but I could have got away with much lower, indeed the Z axis plate which carries the whole Z axis, I have made adjustable so I can raise it for the 4th axis and lower it when I need to get really close to the table, takes ten minutes to change over.
I'd really love to see your first machine though, it's fascinating to see how people have got around problems and even if it didn't work for you, ther might be something in there that could be the spark of an idea for someone else.
Re: the long screws for the 5m axis, did you not consider rack and pinion for this? it would have given you the speed and maintained a good degree of accuracy, certainly the same as you achieved using screws. On a machine that size I think I'd have gone for R&P on at least the X & Y but until I see this monster (hint, hint), who can say what we would have done.
Anyway, thanks for the detailed reply, I like people who dream large :)
Jeff.Nothing is foolproof......to a sufficiently talented fool!
I always find it's the third time of doing something that it becomes right.
As it happens, this will be the 2.5th or 3.5th attempt (depending on how it's counted).
I did two very small-scale mock-ups of the machine - the first was to prove the electrics (pretty trivial), and the second was as an exercise in working out which bits would be the tricky part (the Z-axis, as it happened).
Then, there was the full size machine itself, which started off with standard 10mm threaded bar (I was on a /very/ tight budget to start off with), which were replaced with R&P for speed.
Now I've proven to the workshop owner it's doable, the budget is significantly increased.
As to pictures, I'll post some in a few minutes (just need to check if this is any different to other phpBB systems in that respect).
Okay, wasn't too different.
The first one is a picture of the overall machine - the top frame was made out of 25x50mm box section (too small ["It'll be strong enough", I was told]), and the legs and table top from 50x50mm box section.
The second picture is of the X-axis carriage. It used the frame itself as the track, and spherical rod-ends that could be adjusted to clamp bearings in place, and allow any potential twist in the X-frame to be ironed out.
The third picture is a model that I cut out of insulation foam (extruded polystyrene) of the car that I'm designing using the machine. Unfortunately, I leant on the table causing it to flex, so there's a few contours that are a little out as a result. Unfortunately, I didn't realise I'd had that effect until I'd done it a few times, and even after then, I kept forgetting :)
The windscreen's been sanded a bit, but the rest of the body is as expected.
One thing I didn't mention was that I also built a simulator for the machine - it allowed me to develop the STL format file (Stereo Lithography - that my modeling program can export) to my own custom format for the controller I built at home. It simulated the whole machine, and allowed me to check for tool interference without:
- Cutting any foam
- Driving 70 miles to try it
The output of the simulator is identical to what was produced (ignoring the leaning on the table bits).
By the way, I'd really like to thank Gary from Zapp Automation for his help so far! Top notch.