[Vilyoo]

I will leave them empty

[peteeng]

Hi Vily - If any tool is not trammed then one side is higher than the other and will cut a scallop into the surface. The bigger the tool the bigger the error vs the same off axis amount. All tools and Z have to be trammed correctly to get a flat cut. Especially big ones. Peter

[peteeng]

Hi Vily - I find using std SHS and RHS to be a limiting approach. But if your going to use it, select the biggest heaviest sections possible or combinations to create the biggest geometry possible. Its easier to forget about weight initially, design the UBER humungous machine then lightweight it vs taking an incremental step up think, step up think approach. Takes too long to get to the answer thru the many iterations. Think BIG...Peter

[joeavaerage]

Hi,
the thickest RHS and SHS sections I've seen listed are 16mm. They are not common. New Zealand steel suppliers do list them but on indent from overseas.
No idea how much you'd have to order to get some.

200x200x16, 250x250x16,300x300x16. 250x150x16,300x200x16,350x250x16. They are all 8m as standard length.

You might try some of the big architectral steel companies, they might have some bits left over.

Craig

[Vilyoo]

There is only 400x200x16 in my city, and the others are very big.

Check this steel beam, my local supplier have it available. HEM200

[peteeng]

So Vily - Do you intend stress relieving the weldments? This is best practice (and least risk for being able to machine the lands flat) and how are you going to machine the lands for the rails? The sections themselves will not be within the required tolerance to achieve your machine tolerance... Then there is the fabrication distortion to deal with, TSR solves that... Peter

[Vilyoo]

I will stress relief only Z and xz plate, other part's are too big. I can try vibration stress relief on the long parts. I will have the lands for the rails machined in a local company.

[peteeng]

Ok - sounds like a good plan. Peter

[Vilyoo]

Im going to replace all 300x200x10 tubes to 2x 150x200x10 in parallel + sheet 120x15mm 1 on each side. What do u think?

[peteeng]

Hi Vily - The external geometry is the same so why do this? Your only gain is in the two straps on the sides? Peter

[joeavaerage]

Hi,
if that improves the stiffness then yes that is a good move.

You need to strike some kind of balance. In a gantry machine it is usually the gantry closely followed by the Z axis that give the most trouble. The gantry must resist torsion applied by the Z axis/spindle.
What is the point for instance in making say the table uber stiff but only to leave the gantry overly compliant? The parts made will be marred by the compliance of the gantry. Or lets say you make a great job of
the gantry but the Z axis is overly compliant....the parts will suffer in the same way.

I suggest start your design with the gantry, and basically make it as stiff as you can afford, both weight and cost. Thereafter the remainder of the machine can be designed to get the best from that particular gantry.
Each part, say the Z axis or the rail upstands, need be stiff enough that they do not contribute overly to compliance.....but neither is there any thing to be gained by making them uber stiff and expensive.

My machine is a C column design. In such designs the column is most problematic followed by the headstock. To achieve the best result the design needs be balanced.
The process is design the column to be as stiff as you can within the limitations of weight, size and budget, and then design the rest of the machine to suit.

Craig

[Vilyoo]

changed Z to 20mm steel plates, gantry is now 2x 250x150x12.5mm tubes. The difference is very big from the first post.

[Vilyoo]

Hi Vily - The external geometry is the same so why do this? Your only gain is in the two straps on the sides? Peter

, So bigger and hollow is better.. then how about 400x200x16mm ?

[joeavaerage]

Hi,
bigger is better. The first moment of area is very dependent (to the square? or maybe the cube?) on the outer dimensions. Increasing the size will easily make the structure stiffer even without making
the wall thicker.

Craig

[peteeng]

Hi Vily - Yes the 400x200x16 is considerably stiffer globally then the 300x200 arrangement, plus the 16mm is stiffer locally and it has some meat in it for a machining allowance.... Since the big continuous loads are in the XY plane use the 400mm in the flat so its 400mm wide and 200mm high. In regard to the I beam you put up, I beams and channels have nearly zero torsional stiffness. So they are OK to use in the table but not for a gantry. Peter

[joeavaerage]

Hi,
the advantage of 16mm wall is that its thick enough that you can drill and tap with decent size bolts. 10mm or even 12mm wall does not really allow you to drill into the end
with anything but small screws, say 6mm. With 16mm you can easily drill and tap for 8mm and with care 10mm screws. With any sort of luck you maybe able to avoid welding and that
would save on stress relief.

Of course, 16mm wall is brutally heavy and with that comes expense it does however set you up for a very rigid design, and that rigidity will be paying dividends long after you've forgotten the bill.
If you have access to 400 x200x16....get it.

Craig

[Vilyoo]

Good news, i found a supplier in my city with all different options for tubes. I think im going with 350x250x16mm

[joeavaerage]

Hi,
that's good news.

16mm wall sections are very much the exception here in New Zealand which make it a very expensive option to procure.

Craig