[Boomslang]

Nice! Good to see the cyclone design validated.
Now start building!!! If anything flex in the end, you can just add some more metal. Would love to see some deformation simulations of the cnc with ie. 100kg at the spindle end.

[MaxTheModder]

thx boomslang!
I did already try to simulate the flex - but solidworks stikes on meshing the model for deformation simulation. (does anyone from you know how I can avoid that?).
At the moment the whole gantry weights 72 Kg. I dont know if the steppers will take more when more metal is atteched to reduce flex. I will try on simulation the deformation and according to the results I will adjust.
I think I will have to make a little compromise in rigidy and acceleration.
But I think 1000N will be not a big problem for the design - maybe around 0.05-0.2mm flex in force direction.
Anyway - I will check it with a simulation.
Does anyone of you know with what forces I will have to deal, when I mill aluminium? around 1000N? or 10000? I have absolutely no plan...

[Boomslang]

I went with 100kg load, but that's also just 'thumb suck....
Try making the mesh smaller in the simulation, exclude parts that will not affect the results, If you have a round bearing face running on a flat surface it will have an infinitely small contact point…. check that you don’t have any of those contact points.

[MaxTheModder]

greez all cnc enthusiasts!

finally a picked up my work again...

I was a little dismotivated since my last post - I am not happy with my design and I found out that there are many minor big aspects to change.
If I would spend such a lot money - I would be really frustrated, if it would not work as I wanted. Of course I cant make the "perfect cnc" but I would like to realize something that is at least worth the money.

First I thought that I bet with 20mm aluminium gantry sides on the wrong horse.
The other solution (a LOT cheaper) was to replace all aluminium parts with steel.
It is a lot stiffer then AL but has 3 times more weight -
So I reduced the thickness to 10mm.
That was still to heavy (it was 40kg or so more that the AL design) so I decided to customize the steel gantry sides (saves me about 8 Kg and I loose only 15% of stiffness)
I simulated a lot around with SW and the design below is best I was able to:





But then I told the story a friend and he meant that the steel would deform after some time - AL would not
This was already known to me but I neglected it so far that I almost forgot that. He opened my mind - my design is somekind a fail cunstruction to my needs...
It is too good for very basic mill jobs and too worse for milling Metals and that stuff. So then I thought why should I spent that money for something that is not good enought for that what it was almost "designed" to.

I need to recunstruct - maybe I gonna get new rails for free (4m or longer) but that would be redicolus to "build" a cnc around as I did before (rails in the design already exsist.

PLEASE tell me what you think - your opinion about my problem - what I could make better?

My thoughts so far:

I gonna make direct drive
Maybe the ballscrews on the outside (X)
make the table maybe below the X- Axis (hence shorter gantry side)
Maybe smaller and more rigid then

?Redesign the Z-Axis? :
What do you think about my Z- Design?

- I had to place the stepper with a belt because there was no room when driving it directly
- as in my first post I explained that that desing saves me a lot of weight and gives me a lot more rigidy - big con is that timing belt (deflection......)


SO all in all I gonna need an "idea-reboot" - I think too much inside my box - which has got very sturdy since the last year designing my cnc

grz max

[PaulRowntree]

Hi Max;
I haven't followed this thread, so excuse me if I have missed something important. I would say that making cutouts on the gantry sides to save weight is interesting, but perhaps a bit ahead of the game. First off, 20 mm Al should be lots strong. I expect that the weakness will be the limited width of contact with the bearing blocks, not the sides themselves. If your really need to trim weight, then go thinner and add substantial stringers to transform the side plates into side boxes (similar but stronger than the channel piece that you show on each side). Two pieces of 20mm thick, 75mm high Al mounted on edge to the sides would make the sides extremely rigid, with minimal additional weight. Place them parallel to the cool rakish angle of the sides, as far apart as possible, and you should be good to go.

I see only a single vertical series of bolts holding the gantry rail; to each side. THis is very weak, and does nothing to prevent racking/twisting of the gantry. Can you get two lines of bolts on each end, spaced as far apart as possible? Again, look for ways to turn plates into box-like structures. For the same strength, the box will be much lighter.

[JerryBurks]

When doing material stiffness (modulus) and thickness comparisons keep in mind that for the same material the rigidity of a plate goes with the 3rd power of the thickness. That means if you go from 20mm aluminum to 10mm steel the plate will flex much more although the steel itself may be 3 times stiffer (and heavier). Anyway, best is what Paul suggests building a box structure (for bending and twisting) or a sandwich/I-beam (for bending only).

As I learned with my build the position of the Y-screw is quite important. The closer you get the to tool tip the better you are off concerning the load leverage on your Y-linear bearing. I.e. if your design allows move the screw down as far as possible.

[Villacherman]

Great renderings and info, I've been reading an lurking, but I'm now starting my build.

What add-on are you using with
Solidworks to do 1) the photo-like renderings and 2) the stress simulations.

Thanks!

[RomanLini]

Thanks for posting the cylcone simulation.

This is not really a suggestion but a guess, what would happen if you made the cyclone cylinder larger diameter (+30%) and leave the outlet tube the same size, and a much less pointy cone on the bottom? Say a cone from full width to 40% width? I have a feeling you might be able to get even lower airspeeds at the bottom.

[MaxTheModder]

thanks for your suggestions!

Now I am going to think about a box like structure on the gantry-sides. Maybe I choose 10mm thick AL and add AL-extrusions on the outside

Something completely different I have been thinking last time, is to lower the whole working surface, to let the X-Rails go almost on the height of the Y-Axis.
Then I should not worry any longer about the sides. (But it makes is more expensive)

I see only a single vertical series of bolts holding the gantry rail; to each side. THis is very weak, and does nothing to prevent racking/twisting of the gantry. Can you get two lines of bolts on each end, spaced as far apart as possible

I dont really know what you mean Paul. Do you mean the screws from the Y-extrusions to the gantry side?

If so, I see that there is a problem - but in design it was meant by me, that the lower extrusion takes all (almost) the force in X and the upper in Z direction.

Thanks for the suggestion to place the Y-Screw lower - I tought it would be better to place it exacly between the rails, to avoid shattering - but when cutting, the central Y- force goes to the bottom and there should be the screw, as you said. I gonna do that -thx

@Villacherman:

I am using SW Premium 2010 (Student) there is already a stress simulatioan plugin included and the rendering are done with PhotoView360 (come also with SW Premium)

@Romanlini:

You may be correct, but making the main diameter bigger, results to a lower vilocity along the wall.
I will focus on the cycolne more when finshed the CNC-design - I will post the other simulations for you, if you want. Maybe I will make a different cyclone design - lets see when its time for

PLEASE MORE SUGGESTIONS!!! (dont be shy)

grz MX

[PaulRowntree]

thanks for your suggestions!

I dont really know what you mean Paul. Do you mean the screws from the Y-extrusions to the gantry side?

If so, I see that there is a problem - but in design it was meant by me, that the lower extrusion takes all (almost) the force in X and the upper in Z direction.

I now see what you are doing there on the gantry Max, with two Al extrusions at 90 degrees to each other, to handle the racking and twisting separately. In effect, you are boxing the gantry. That is an interesting approach, and I don't remember seeing it before. Tie the two pieces together on the backside with a triangular right-angle gusset at each end, and it will be very stiff. Cool!