the build is slowly starting.
poor picture so far, but it gives a decent idea.
It's a bolt on jobby. still need a few brackets to make it sturdier. The X axis beams are 3mm thick steel. damn heavy...and damn bent. i need to find new ones as they have about 1mm in them.
Questionable Support! stef_d,
What do you plan on cutting with your machine? IMO, your long X-axis rail support "beams" should be given additional support. Perhaps run another "beam" parallel to the main one and tie them together with uprights every 6" to 10". I realize you are using "Supported Rails", however if you plan on routing hardwoods or aluminium (even with shallow DOC), you will need additional stiffness/rigidity along that X-axis (especialy if your gantry has any weight to it). As it is early in your build it shouldn't be too difficult to add it now, as oppossed to trying to add it in the later stages.
So, what do you plan on cutting/making with your machine? Please give us all the vital statistics (working envelope, motor sizes, electronics package, router/spindle, etc.).
Good luck with your build. Thanks for posting, axiously awaiting additional updates with pics!
Randy,
I may not be good....
But I am S L O W!!
The goal is to cut wood, and maybe a little aluminium.
I did think of adding more support. but trust me, those beams are stiff enough.
I calculated the deflection with a 100Kg weight in the middle. and that is just enough to make if move something like 0.05mm.
It's 50x50 3mm thick steel. weights about 20Kg a piece. (shame that they are bend)
they will have to be replaced anyway.
Hi Stef,Originally Posted by stef_d
Which way way are they bent,?
If i can explain my self properly i'll try and explain a liitle problem i had with bent or not straight rails, and the solution.!!
My top rail was bent along it's length, so viewing from the side of my machine the top rail was bent if you imagine in an extreme way like a banana.
If viewing from the front it's bent then it doesn,t matter because you position the rails on the topface so you can adjust for this and infact you should work from a datum on one side so then you can be sure they are perfectly parallel. I used a precision straight edge and tak welded some washers to the rail that where butted against the straight edge, this was the used as a datum line and the rail on this side aligned and bolted down and positioned against these washers. The reason i used washers was so the carriages could pass over them and also it meant i didnt have to find a perfectly straight piece of steel, just put washers every 200mm or so.
Then what you do is position the other rail on the opposite top rail and use a device to get them as parallel as possible then very losely bolt them down then run your gantry up and down which will self align them and slowly working your way up and down tighten and adjust untill as smooth as possible.
Right the cure for the bent toprail was incorporated with my adjustable bed but could still be used without it.? . . . The threaded rod that my bed raise's up n down on also is used to pull and adjust the toprail as close to flat as possible and you'll be surprised just how close it can get it, then i ground and filed, scraped any high spots and filled any low spots with braze untill i got them as close to perfectly flat as possible using the precision straight edge then with a few blades from feeler gauges i shimmed any places that needed them.
The threaded rod also adds support from deflection.!!
If you want to know more you have my number give me ring or email me.
I dont think you will find a piece of steel that is perfectly flat and if you do then the minute you try to weld it chances are it will move.
The welding at best will need to be balanced out and done in small burst's and then move to opposite side to balance the stress's.
Cheers
Dean.
PS: The wine was lovely i've just drunk the last bottle tonight.
OK, forget the welding i just spotted your bolting every thing together.
One other thing i would suggest and only thing with my machine i would do differantly is to fill the hollow box section where ever i could with something to dampen vibration/noise.
It doesn't affect the cutt as far as i can see so far on both the wood or aluminium that i,ve cutt with it, and i,ve cutt thick-ish Alu 20mm 6082 t6.!!
That said i did take very shallow cutts but with a high feed rate, it worked a treat and i'm very very pleased with it.
stef,
I think your calculations are off. I happened to have done calculations for 3x3" .120 A36 tube recently. Yours is smaller (2x2x.120). Under its own weight I get 0.018" deflection in the middle of an 8" span. Your ~1mm is in the same ballpark. Stick a heavy gantry in the middle and you'll probably get 1/8" or more deflection.
Dont get me wrong this is not aimed at you directly, but it makes me laugh when i see all you people wasting time doing stress calculations that dont often work out like the computer predicts because other factors that come into play that the simulation doesn't take into account or you dont input into the equation.
Only to cutt wood 90% of the time which alone will deflect more than any part of your machine ever would the minute you put a cutter near it.
Come on get a grip, so long as the materials used are upto the job(and if you need a computer to tell you that then you shouldnt be considering building a cnc machine your self.!!) then unless your talking milling machine tolerances just get on with the design and build the bloody thing.!! . . .IMO
Dean,
Glad you liked the wine !
for the beams, they are bent both ways (side to side and up and down). they had been welded before. New ones are cheap, so it's no big deal. only the trouble of drilling 32 holes in each once again !
I tested them by resting one on top of the other, rolling surface against rolling surface. they diverge by about 1mm.. i tried the same thing with new tubes, and those are much flatter. no gap in between.
for the calculation, i checked again: here are my figures:
Iz=1 041 666 mm4 (50x50x3mm)
L=2000mm
Young's = 210 000 MPa (steel)
and force =500N (i used half of a 100Kg Gantry. mine is probably closer to 30Kg (it's aluminium profiles). placed at a single point, in the middle of the beam.
Total flex is 0.095mm. worse case.
using more realistic figures (30Kg gantry, on two cariages, 200mm apart), it's more like 0.028mm.
If i calculate the flex from the weight of the unsupported beam itself, using 0.1N/mm as the linear force of its own weight, i end up at 0.019mm
add the two, and we are under 0.05mm.
not too bad.
where dont we agree ?
Looking at the base last night, i wasnt too happy.
i need to go through another design phase before it's too late.
I am particularly unhappy about the top supporting beam connection to the legs. they will pull on the feet and open them up. Those beams need to rest on top of the feet, not hang on the side.
Also, the way the base structure is setup. i can do better.
I i brase some corner piece to the 40x40 beams and then bolt those on the feet, it will be neater.
This is better, i hope:
I also need to put in some adjustement threads, like you have setup up, Dean.
do you have a good picture of what yours looks like ? I think i know what you mean, but a picture speaks a thousand words !
back to the drawing board for a bit..
Actually Dean, I looked at your build thread again,
I am just a little unsure: your cutting bed is only supported by the 6 threads ?
the rail suport beams, at the top, are welded to the feet, but can still be "lifted" flat using the same rods.. is that right ?
Yes your right about six threads but needn't have any fears because i can tell you it's absolutly rock solid and doesn't deflect or move in the slightest.
Yes i'm sure if you put the figures into a simulation the comp will tell you it moves this much or that much but i,m telling you now from the experience of using it that it does not move in any way that affects the use or accuracy of the machine for what i,m using it for. . . Which is machining woods of all kinds light/heavy hard/soft and drilling/profilling Aluminium plate upto 25mm thickness.
Maybe if i put an engine block smack bang in the middle and tried to surface plane it then it might.? . . . .but it's wasn't built for that so why should it.!!
Remember mostly the the weight is spread over a larger area esp when using wood.!!
No the the rod is bolted to the top rail with a nut eithier side and the lower rail is used as the adjusting point which push/pulls the top rail up/down how ever i want it then this locked off with lock nuts, nothing can move or slide.
The bottom rail does have to be supported from below and if you look you can see the diag brace coming from the legs to support them, which to be honest you would need even if you just sat the bed frame straight on top.
Plus it stiffens up the whole frame.?
It's not the most elegant looking thing and i,m not saying it's the best way to do it but i am telling you it WORKS GOOD and in the real world not virtual.!!
We disagree on two points: 1) Your Izz is 5x too high. I think you computed the Izz of a solid bar and threw in a factor of 2 (I get your answer from 50^3*50/6). The Izz of a solid bar would be 50^3*50/12. Subtract out the "hole" of 44^3*44/12. I get 208492mm^4. If I look up "2x2x.120 inch" in a table I get .486in^4 or 202288mm^4, so I think I've got about the right value.
Also, to get your results I have to use the formula for beam deflection with fixed ends. In other words, the leg of your table would have to be resisting the inward pull caused by the rail sagging. In fact it will just bow in. So instead of Pl^3/192EI you want Pl^3/48EI, which is another factor of 4.
So multiply your results by 20x. I also assumed a heavier gantry.
ahah, yes, you are right.
I miscalculated the Izz.
What i need it to substract the Izz of the void in the middle of the tube, which is 44x44.
all in all giving a new Izz of 624682.
plugging this back in, using 500N as the force (only point force) i end up with 0.158mm.
back to a realistic 150N. and its' down to 0.047mm. (not counting for flexing support and own weight)
To be honest, with this new figure, i will provide added support. but trust me, i can jump up and down on the bar, it will not flex by 1mm. The bend i can see now is not flex. as i said, i can see the bars diverging when one is resting on top of the other, on the floor.
thanks for the input anyway.
Posting Permissions
LinkBack URL
About LinkBacks
