Re: Moving gantry conundrum
Quote:
Originally Posted by
HardDreamer
Hi,....
The usual way of doing this, would be having two ballscrews on the gantry moving axis. This conflicts with my closed loop/DRO design, since this design cannot possibly compensate for two crappy ballscrews at the same time.......
Put a linear encoder on each side, read the actual position, and take the ball screws out of the equation.
Re: Moving gantry conundrum
Hi,
I think your requirements are mutually contradictory......perhaps not impossible.....but not at all easy.
You are correct the milling operation, and milling steel in particular, is by far and a way the most demanding of rigidity of your machine.
May I suggest you look closely, and then look again, at the likes of Grizzly and Rong-Fu drill mills. I'm not suggesting that you buy one necessarily
but that you consider what the manufacturers have found that works. Both these machines are rigid enough for light milling in steel. They are NOT over rigid,
otherwise the manufacturer would reduce the size/weight and cost of their machine....but they have not.
If you are to build a machine that can do light milling in steel then you will need to attain AT LEAST the rigidity that these machines exhibit.
Moving gantry machines are not the easiest approach, the gantry has to be MASSIVE in order to be rigid enough to mill steel.
What you have described about what you want from your machine sounds like two different machines:
1) A gantry style router to do 3D printing and EDM with a generous work envelope....and
2) A conventional bed or knee mill rigid enough for steel.
Trying to combine them into one is going to be hard and likely the compromises forced on you will mean lesser performance
than you hoped for.
Your design task would be VASTLY simplified if you drop the requirement that it do steel.
This may not be what you want to hear......I was told the same thing six-seven years ago when I was designing my mini-mill.....and I
ignored it. I wanted, and eventually built, a very small bed mill using ground ballscrews, HD rectangular linear rails and cars mounted on cast iron
beds and a solid steel 75mm x 75mm column.
My intention was that it was rigid enough to do steel....and it does....but only just. You might think What I described would be any amount rigid enough
but the truth is I can only do steel in very light cuts. My machine has performed admirably in brass, aluminum and plastics, and I use it for isolation
routing of circuit boards daily.
I have used my little mill extensively and its paid for itself many many times over....but for all of that it has limitations and rigidity is the absolute key
requirement.
I am building a new mill, again a bed mill. The travels are 350mm x 350mm x 350mm using C5 ground 32mm doublenut ballscrews, 20mm HD THK linear rails/cars
mounted on 115kg cast iron beds which are in turn mounted on a C frame constructed of 32mm steel plate. Using 750W AC servos on all axes. If you think that
sounds expensive....it is....but that's what necessary for milling steel.
Craig
Re: Moving gantry conundrum
Here's a point of reference for how thick your parts might need to be, can't speak first hand about this machine as it's not mine, but yours will need to be at least this solid: https://www.youtube.com/watch?v=bX2E...el=CNC_channel
If you go far enough down the ballscrew rabbit hole you'll probably come to a similar conclusion of many others here, which is that getting C5 ground ballscrews with proper preload from alibaba/aliexpress is the well worth the extra money (not as much as you might think)
Re: Moving gantry conundrum
Hi Joeavaerage
Thank you for a very comprehensive answer!
I looked at the mills you mentioned, and was actually quite surprised by the reasonable prices, but since I am focusing on a computer controlled machine, they will need a lot of work to get there.
You present a pretty convincing argument, and I will drop the EDM 3D print angle and just go with the steel mill, to not confuse things. On the other I have studied this:
https://www.youtube.com/watch?v=gUFeoDf1YvM&t=421s
And he seems to get very decent results with a moving gantry, but I have not seen examples of cutting steel, I will have to think some more about this.
750W, holy cow!
A lot of food for thought here!
KR
Henrik
Re: Moving gantry conundrum
Hi Catahoula
Actually, the ballscrew I have bought already is C5, but paper is cheap, I will believe it when I have measured it out :-)
KR
henrik
Re: Moving gantry conundrum
Hi HD - At 350mm wide you are at a point where its worthwhile considering a simpler C column design. Depending on what you consider is "cutting steel" the gantry turns out to be huge as others have said. You need to look at commercial machines that do the same thing you envisage and see what size they are. Do not think they are that big for no reason!! If you reduce your Y to 300mm you definitely can go the C column route. Maybe 300x600mm is a better table size...This reduces the issue of multi drives axis drives and reduces the arm length of the spindle. cheers Peter
Re: Moving gantry conundrum
Hi,
you are narrowing your design parameters and that avoids the 'trying to make one machine do multiple jobs', and that is
all to the good.
Quote:
I looked at the mills you mentioned, and was actually quite surprised by the reasonable prices, but since I am focusing on a computer controlled machine, they will need a lot of work to get there.
Yes those machines are reasonable value, and over the years I have used a few of them for projects, all manual control. Used within their limitations they are
great. The reason I mentioned them is because they are what I would call the 'lowest possible denominator' in terms of rigidity. Any less and they would become useless for milling steel.
Note that there are many which have converted to CNC machines of various types.
The simplest approach is to replace the X and Y leadscrews with ballscrews and retain the dovetail ways and gibs. Some users are happy enough that only X and Y axes
are computer controlled while others use computer control over the quill and /or the column. There are quite a few videos on Y-Tube about such conversions....well worth a look,
you may be surprised how effective they can be.
The important point here is that the manufacturers of these drill mills have arrived at a balanced design. The travels, Y in particular, mean that a C column is a viable design
choice. Had they wished much larger Y travel then the C column design becomes over large and expensive to attain the required rigidity, and if they failed to make it sufficiently
rigid they would fail (miserably) in the market place.
You can build a machine that is superbly fast, or smooth, or highest resolution or the prettiest colors (lol) but it will be just so much junk if its not rigid enough for the work you want to do.
Peteeng, who as posted in this thread, has another long running thread about the design of a bench top milling machine rigid enough for steel. I would advise you read through it
and appreciate the design process. Pete has a long history with FEA analysis and so he designs to a target. He has determined that to meet the machining requirements that he
needs rigidity of 20N/um or better in all axes.
https://www.cnczone.com/forums/cnc-d...11248-cnc.html
Pete has determined that he wishes to pursue a composite design.....and his thread is dedicated to that approach.
I on the other hand am a traditionalist and favor cast iron (when I can afford it) and steel frame (when I cant). My machine has somewhat smaller travels which were determined
by what (quality) secondhand or new old stock components I could find, but is considerably bigger (footprint 1300mm x 1100mm) and heavier (approx 1000 kg). I am hoping that I might
achieve 100N/um in all axes but my expertise in FEA, or rather lack of expertise, means that I am not confident in my prediction a-priori, but must measure it a-posteriori.
Where both Pete and I agree is that we both have a target in mind.....and work backwards from there. Pete's professional experience shows bigtime.
My first design, my mini-mill of six-seven years ago, was more a 'build it and see approach' and was certainly not a 'fail' I had still underestimated the rigidity required
for steel. With the money I have sunk into my new build I CANNOT afford a 'fail' and I have adopted this list of priorities:
1) Is it rigid enough for steel and stainless, say 100N/um?
2) Is it rigid enough?
3) See the above
4) Is it well damped?.......very much a secondary consideration to the above......see 1), 2), and 3).
5) Is it fast......you know I don't really care how fast it is so much as I care about 'Is it rigid enough?' see 1), 2), and 3).
6) What's its resolution?......only really matters if 'Is it rigid enough?' see 1), 2) and 3).
You will observe that I am quite fanatic about that because just about any feature of a machine can be improved or tweaked....but not its basic rigidity....its
either built/designed in from scratch or its not there.
Craig
Re: Moving gantry conundrum
hI hd - tO SUMMARIES cRAIGS NOTES. oPPS CAPS LOCK _ The machine rigidity is its DNA. It has to be there at the start for a good machine. The forum is littered with machinists making their mills & routers stiffer. These include Commercial machines and Maker machines. Its taken me many many designs and many years to appreciate just how stiff a steel mill has to be. Every part has to be very stiff and considered carefully, my current thoughts are about the connections and the parts foundations even the shape of the rail support areas is important. Good luck and keep at it... there's huge depth of knowledge and help here...Peter
Re: Moving gantry conundrum
Quote:
Originally Posted by
peteeng
Hi HD - At 350mm wide you are at a point where its worthwhile considering a simpler C column design. Depending on what you consider is "cutting steel" the gantry turns out to be huge as others have said. You need to look at commercial machines that do the same thing you envisage and see what size they are. Do not think they are that big for no reason!! If you reduce your Y to 300mm you definitely can go the C column route. Maybe 300x600mm is a better table size...This reduces the issue of multi drives axis drives and reduces the arm length of the spindle. cheers Peter
Actually a C beam with no moving bed was sort of what I started with, but now I am more focused on a design like this:
https://www.youtube.com/watch?v=rXs4fsPsSGw&t=3s
https://static.wixstatic.com/media/c...plan%20fr.webp
Website:
https://www.flyingfoxcam.com/cnc
If it is stiff enough for routing hardened steel, I guess it is plenty stiff enough for what I want to do.
He is using some pretty expensive parts, which I probably cannot afford, but building a cheaper version and upgrading as necessary will hopefully work.
BTW: I am really impressed with your work, I have not been quite through it yet :-)
KR
Henrik
Re: Moving gantry conundrum
Hi Joeavaerage
I get it: "Rigid" :-)
Thank you for the link, good stuff!
KR
Henrik
Re: Moving gantry conundrum
Hi HD - Notice the size of the linear bearings on the flying fox. They are very big. Are you going to go the real granite route or mineral cast? cheers Peter
Re: Moving gantry conundrum
Hi,
Youngs Modulus of granite is about 50GPa, cast iron in the range 100-150GPA and mild steel about 200GPa.
So what's the attraction about granite? If a component made out of steel is one inch thick, then you'll need closer to 2 inch thick granite to get the same stiffness....why?
Steel and cast iron you can cut, drill, tap and machine whereas granite requires stone mason tools?????
There is a very good reason that steel is the pre-eminent material for engineering purposes.....its strength and stiffness relative to its cost is fantastic.
Craig
Re: Moving gantry conundrum
Quote:
Originally Posted by
catahoula
Beautiful, but I am not sure my wife would approve :-)
KR
Henrik
Re: Moving gantry conundrum
Hi Craig
Some solid arguments.
But there are other factors as well:
"
Several advantages of granite over steel became evident.
Granite is harder, although more brittle and subject to chipping.
You can lap Granite to much greater flatness and faster than steel.
Granite also has the desirable property of a lower thermal expansion compared to steel.
"
from: https://www.dimensionalgauge.com/201...e-calibration/
He is discussing surface plates and not CNC machines, I know.
Also granite dampens oscillations better that steel.
Besides I got a suitable block for around 70$! I am not sure that it is good enough, granite is not just granite!
I am not done thinking, I will make a drawing when I figure out how to use a 3D CAD program :-)
KR
Henrik
Re: Moving gantry conundrum
Hi,
Quote:
You can lap Granite to much greater flatness and faster than steel.
Steel and cast iron lap every bit as well. In the century or so prior to the uptake of granite for metrology purposes all precision plates and angles were cast iron,
until steel manufacture became commonplace, and cast iron was steadily replaced by steel.
Just as a matter of interest have you seen anyone lap granite? Did you know that cast iron 'Master Plates' are used as laps? Three master plates constitute a set
and can be made arbitrarily perfectly flat using the method first proposed by Whitworth in the 1830's.
Are you seriously considering lapping/cutting/grinding granite?
In fact lapping any hard substance to any great degree is time consuming, granite no less so than steel.
Quote:
Granite also has the desirable property of a lower thermal expansion compared to steel.
True, but are you planning on a temperature controlled machine? Sounds way more than a hobbyists pursuit.
Quote:
Also granite dampens oscillations better that steel.
True, but spaghetti dampens vibrations better than granite....would you build a CNC machine out of it? The truth is that stiffness beats damping hands down every time.
There is a reason that damping in fourth on my list of objectives and stiffness (or rigidity) are the top three. The rule is make it rigid.....then worry about damping.
May I suggest an experiment. Put two equal height blocks under either edge of your piece of granite then put a 10kg weight (100N) in between the two and
measure the deflection of the granite. If you follow peteengs argument you will know that it needs to deflect CONSIDERABLY LESS than 5um in order to
be stiff enough for milling steel. Now find a piece of steel, say an inch thick and the same width and repeat the experiment. Which deflects more? The stiffest is the best for machine construction.
Quote:
Besides I got a suitable block for around 70$! I am not sure that it is good enough, granite is not just granite!
Does the fact that 'its cheap' make it any more suitable for CNC use?. Remember you will invest hundreds if not thousands of dollars and hundreds of hours into this machine,
$70 here or there is small beer in comparison to 'does it work as I had hoped'.
As an example, when I built my mini-mill I bought three 100kg cast iron elevator weights from the scrap metal dealer. They were cheap. I proceeded to machine axis beds out of them.
Of course the cast quality was poor, after all a 'weight' only has to be heavy. There were a few inclusions in the cast and a hard rime on the outside. They did work out OK in the end
and I am still using them today....but I vowed that the 'next time I will get axis beds cast for me'. The compromises that 'cheap' force on you live the life of the machine,
and when it comes to basic structure that is often fixed and beyond any reasonable means of repair or improvement after the fact.
Are you a stone mason, or have access to diamond tooling and machines? That sort of equipment is even harder to find than tools and machinery for steel.
Craig
Re: Moving gantry conundrum
Quote:
Originally Posted by
joeavaerage
Hi,
Steel and cast iron lap every bit as well. In the century or so prior to the uptake of granite for metrology purposes all precision plates and angles were cast iron,
until steel manufacture became commonplace, and cast iron was steadily replaced by steel.
Just as a matter of interest have you seen anyone lap granite? Did you know that cast iron 'Master Plates' are used as laps? Three master plates constitute a set
and can be made arbitrarily perfectly flat using the method first proposed by Whitworth in the 1830's.
Are you seriously considering lapping/cutting/grinding granite?
In fact lapping any hard substance to any great degree is time consuming, granite no less so than steel.
True, but are you planning on a temperature controlled machine? Sounds way more than a hobbyists pursuit.
True, but spaghetti dampens vibrations better than granite....would you build a CNC machine out of it? The truth is that stiffness beats damping hands down every time.
There is a reason that damping in fourth on my list of objectives and stiffness (or rigidity) are the top three. The rule is make it rigid.....then worry about damping.
May I suggest an experiment. Put two equal height blocks under either edge of your piece of granite then put a 10kg weight (100N) in between the two and
measure the deflection of the granite. If you follow peteengs argument you will know that it needs to deflect CONSIDERABLY LESS than 5um in order to
be stiff enough for milling steel. Now find a piece of steel, say an inch thick and the same width and repeat the experiment. Which deflects more? The stiffest is the best for machine construction.
Does the fact that 'its cheap' make it any more suitable for CNC use?. Remember you will invest hundreds if not thousands of dollars and hundreds of hours into this machine,
$70 here or there is small beer in comparison to 'does it work as I had hoped'.
As an example, when I built my mini-mill I bought three 100kg cast iron elevator weights from the scrap metal dealer. They were cheap. I proceeded to machine axis beds out of them.
Of course the cast quality was poor, after all a 'weight' only has to be heavy. There were a few inclusions in the cast and a hard rime on the outside. They did work out OK in the end
and I am still using them today....but I vowed that the 'next time I will get axis beds cast for me'. The compromises that 'cheap' force on you live the life of the machine,
and when it comes to basic structure that is often fixed and beyond any reasonable means of repair or improvement after the fact.
Are you a stone mason, or have access to diamond tooling and machines? That sort of equipment is even harder to find than tools and machinery for steel.
Craig
Again, very convincing arguments. Still I do have a quite a picture on how to go about this, which fits the material quite well. I have a angle grinder with a diamond wheel which cuts granite very neatly, and diamond cups and drill are not unreasonable.
Try have a look at this:
https://www.youtube.com/watch?v=gUFeoDf1YvM
and
https://www.youtube.com/watch?v=9yBWrzBjW5w&t=338s
The drill stand he is using is around 300$
And yes, I know that steel is used for lapping granite :-)
BTW: I have been looking at Peteeng's work I am completely sold on the lifting gantry concept, which raises a whole new bunch of interesting problems.
But, I really need to get a drawing put together to explain what I mean. This weekend I will get it done!
More to come.
KR
Henrik
Re: Moving gantry conundrum
Hi,
yes I've already seen those videos.
I have three steel saddle plates 400mm x 400mm x 20mm. I had them ground on a surface grinder, they came back flat and parallel to within 1-2um.
It cost me $20 and a plate of scones for the guys, I dropped them off at 11am and picked them up an 3pm. Is that cheaper, easier and quicker than lapping
granite?
Craig
Re: Moving gantry conundrum
I think, it was a great solution to watch the video on YouTube.
Well, alternatively, I can offer to cooperate with a company that has been in this business for a long time. For example, you can look at https://engre.co/ there are many companies to choose from or still try to do it yourself.
