cnc steel frame

[peteeng]

Hi Ben - Coming along nicely. I didn't know that you want to use the large trans tubes as the feet. Its good you have moved along to the gantry. Now for the Z axis then around the cycle another couple of times. Do you have access to a good machine shop? Peter

[bensaad]

hi peter
-Yes, I want to use the big red tube as legs for the machine to facilitate the process of assembling the machine ,because
those tubes are heavy and will not move during assembly and use less bolting
-Yes I have access to machine shop, sometimes I go to a friend of mine who has a shop that sells these machines to get som inspiration

[jaguar36]

Have you considered how much all of this is going to weigh? And how much all of that steel is going to cost?

You've got an awful lot of steel (and therefore dollars) just holding the machine up. You've also got your bed only what 500-600mm off the floor? That's going to make working with it pretty uncomfortable. I'd redesign it with some legs, minimizing the amount of steel holding the machine up and maximizing the amount that is helping stiffening the load path from the cutter to the material.

[joeavaerage]

Hi,
I agree with Jaguar, steel is just the best and stiffest material, so use it, and heaps of it, when you need to make something rigid, but use something lighter and cheaper like wood for anything that does not
contribute to the rigidity of the machine.

I urge you to have a look at Jaguar's build thread, he is bolting stuff together, and it works.

Craig

[joeavaerage]

Hi,
you asked earlier about how you'd go about bolting SHS sections together.

One of the advantages of the thick wall 100 x 100 x 9 that I recommend is that you can drill and tap and the wall is thick enough to reach full torque without fear of stripping
the thread.

The other recommendation I would make is that all the SHS sections be put in a surface grinder and 'skimmed flat' BEFORE they are assembled. This would mean finding
a large format surface grinder, but they are around. It would also mean that all the members would be flat and of uniform thickness, and could thus be bolted together with a good prospect
of forming a flat and uniform structure.

Bolting would probably require several gussets around the corners to retain squareness, however as soon as you screw the spoil board down it would do the same job.

Craig

[peteeng]

Hi - Bolting or screwing members (or machine parts such as rails and cars) together means the connection is relying on friction to transfer strain thru the connection. Friction relies on the preload of the connection creating a strong normal force. If the surfaces are not flat, then the bolt load is shared between the contact forces and the forces needed to deflect the local area. The deflection forces do not contribute to the friction. So yes its important that surfaces are flat. But if not flat but close you can use epoxy to bond the connection together (sometimes called setting the joint). This can be done on connections that you intend not to come apart.... but they can be pulled apart using a heat gun to soften the epoxy if needed. Bolting is a good solution to many structures and rivets are bolts as well. Many hot rivetted structures have stood the test of time.. High strength bolts can produce higher friction forces than hot rivets so are in many ways better then rivets... But hot set rivets look cool Peter

If your interested in aluminium construction the co-efficient of friction typically used for steel connections is 0.3. The typical for aluminium is 0.6 so the connection is much stronger in al using the same number of screws.... One of my prior jobs was testing towbars for static and fatigue performance. We found that using large backing plates vs using washers resulted in transfer of bigger loads and better fatigue performance, CNC Machines are not concerned with fatigue but I'd use backing plates wherever possible to improve local stiffness of the conenction...

[joeavaerage]

Hi,
that is why I suggested that OP grind the members flat (top and bottom) before being assembled. That would ensure a good fit between members, and at 9mm wall thickness
you could really lean on some 10mm or 12mm cap screws without stripping the thread. Aside from which I think surface grinding a whole bunch of longish (say 2m) lengths
of 100 x 100 x 9 all adjacent one another would be a very cost effective way of producing uniformly flat and uniform thickness members.

Remember a very successful aircraft, the Hurricane, was a tubular structure all bolted together with crush tubes, doublers and gussets. They had a great reputation for surviving combat
damage.

A well designed an executed bolted structure could every bit as good as a welded structure.

Quite frankly I'd make a timber frame to hold it off the floor. Why pay big dollars for steel when you can use timber? The only steel I'd use would be as Jaguar described it
'providing stiffness in the load path between the material and the cutting tool'.

Craig

[peteeng]

Hi Ben - Now you have moved on a bit and are looking at the R&P arrangement you need to consider what speed range you want from the machine. Normally an R&P has a 3:1 belt or 5:1 gearbox on the drive. This is to lose some speed from the arrangement and gain some torque. Last time I went through this I think I settled on 5:1 precision gearboxes. A 3Nm N23 motor will produce 173kgf using a 10mm pitch ballscrew. Using a 5:1 gear means a 3Nm motor produces 15Nm and using a 44mm pitch pulley results in a force of 681N or 70kgf. At 1000rpm (top speed for a stepper) this gives one rotation = 138mm or at 1000rpm is 138x1000/5= 27600mm/min which is quite rapid as a top speed. Use a 10:1 and you get 140kgf at 13800mm/min. These are still fast at 500rpm which will be the sort of max cutting speed you'll need at 6m/min... I've gone through this quick so check my math... Pulley systems get real big after 3:1. Peter

PRO Rack and Pinion Drive, NEMA 23 | Avid CNC

PRO Rack and Pinion Drive, NEMA 34 | Avid CNC

Nema 23 Stepper Motor L=76mm Gear Ratio 10:1 High Precision Planetary Gearbox - 23HS30-2804S-HG10|STEPPERONLINE (omc-stepperonline.com)

a direct drive with 44mm pitch pulley would drive at 1rev = 138mm so 1000rpm = 138m/min so say 69m/min rapid at 500rpm.... My first router had a drive belt drive and I regularly finished timber moulds at 25m/min very very fast....

image - A small router I built and currently test things on with a N17 motor and a 10:1 gearbox on the X axis. It would tear your arm off if you got caught in it... This is now testing a new 6 axis controller I want to use on a 5 axis machine I'm designing. The column and saddle are steel and torch brazed (Tobin Bronze) and it uses a 10mm AT5 belt.

[bensaad]

hi and thanks you all
-bolting the frame togother,and surface's grinding all SHS sections
i saw some industial machine weld the frame togother then weld a thick steel bar in the side to mount rail and rack as showen in picture.
i think this solusion is great because the only part would be flat machined and aligned is this bar
Does this work for a diy cnc?

-for the gearbox reduction i want to use the avid cnc r&p drive

-and for the machine legs could i use H beam steel instead beacause Because it is available and cheap,or whats you means by timber craig?

[peteeng]

Hi Ben - 1) If you weld such a structure you will have to stress relieve it. It will change shape during welding and during machining if not stress relieved. Craig means to build the "bench" from timber but the structural members in steel. Many people build successful machine bases and benches from timber or aluminium. I've built entire machines from plywood. There is a German company that builds much larger commercial machines than yours from plywood. All materials are viable once you learn how to get the best from each one. 2) The avid drive is a good hobby grade drive. Your machine is potentially at a level well above avid. You can do better than the Avid drive. Machines need to be consistent and parts need to be commensurate with their purpose and overall quality to make a good machine. May as well buy an Avid kit if your going to buy an Avid drive IMHO. Will save you a year of work and you'll get support and a guarantee... Peter

Re- I beams, universal beams and channel. All are valid if used properly. Are there laser cutters near you? You can have parts laser cut and bent. Then you braze them together. This allows you to optimize the geometry and local thicknesses. Most of my machines have laser cut and bent parts that are bolted together, then I am not limited to std geometry.

I'll be heavy handed on the bed design in the image. The heavy plate is attached to an area of the RHS that has nothing behind it. That plate will twist the side of the RHS locally as loads are applied. Its local stiffness does not match the frames global stiffness ie not a good design.

[bensaad]

hi peter and thanks
You convinced me, and I also couldn't find the steel in the size I designed the frame with. The supplier told me they have stopped manufacturing it entirely.
Currently available size is 120 and below with wall thickness of 5mm and less
I am now faced with two options
-120x5mm or 100x4mm square tube in the load path and maybe 120X60X3mm or 100x50x3mm rectangular tube as cross members part

i redesigned the machine as the picture attached, less steel lightweight and cut down the cost too,but if i use 120x5mm
should i add a thick flat bar to bolt the rail on ,or 5mm thick would be enough?
the machine size is 2750x1500mm
the legs is 625mm tall

[peteeng]

Hi Ben - Move the legs inwards to decrease the beds span or add a middle leg and a middle longitudinal. It may seem stiff but it will bounce a bit in the middle as drawn due to the gantry moving around and tool loads. Is the frame welded or brazed? When it comes to local loads from bearings and racks, thicker is better. The cost of the material is tiny to the project and the machining cost is the same so go thicker.

As a guide if the part is not at a "hard" point ie a place where loads are coming in and out then it can be thin. The legs and frame rails etc make 3mm or thinner. The parts where the bearing rails are and car mounts make very thick. Many machines use light "foundations" where loads come in and out and they locally deflect due to shear deflection. A thin long beam is fine in global stiffness but will have poor local stiffness. Peter

your 1500mm maybe a bit narrow for a full sheet machine. You'll need at least 100mm perimeter for clamping and access so the bed is 1220+200=1420mm then there's the width of the drive say 100mm? so it would be 1620mm? say 1600mm? If space is not a problem then be generous with your sizes. A big machine needs to be big...

[bensaad]

Hi Peter and thank you
The frame will be brazed.
As I understood you I should use 120x5mm square tube and no need to add the extra flat bar,also add two legs in the middle with another longitudinal.
Is that's right

[joeavaerage]

Hi bensaad,
as a rule of thumb the thickness of the wall should be the same as the diameter of the thread.

Thus with a 5mm wall thickness you'd get a pretty good and full strength thread for a 5mm screw, 6mm not so much, 8mm pretty marginal, 10mm or bigger the thread
is just too understrength to be useful.

With 5mm wall thickness you are going to need doublers for any screw over 6mm. Is this want you want?

Craig

[peteeng]

Hi Ben - Is the flat bar the place you are going to mount the rails and rack? if so yes; as this needs final machining or do you intend to machine the SHS directly? You do intend to machine the bearing and rack lands? Peter

[bensaad]

With 5mm wall thickness you are going to need doublers for any screw over 6mm. Is this want you want?

HI CRAIG and thank
i have HGR20 linear i think it need m5 bolt so 5mm tube wall thikness is enough?

[bensaad]

hi peter and thanks
yes the flat bar will be used to mount the rails and rack.
but i asked if using 5mm thick tube, no need to use an additional flat bar,all I have to do is machining the SHS directly where the rail and rack will be installed

[joeavaerage]

Hi bensaad,
once you start machining the SHS you are going to lose wall thickness in a hurry, and then it starts to look pretty marginal. Were you starting with 9mm wall and ended up with 8.5mm or even 8mm
then you'd still be fine, but becomes increasingly marginal as the wall thickness reduces.

Should for instance the wall thickness reduce to 4mm-4.5mm after machining then the torque that can be applied to the screws needs be limited for fear of stripping the thread. Also when the load comes
onto the rail from the car the underlying steel is going to deflect. The thicker the section the less deflection.

I can well understand the drive to reduce the section thickness to reduce cost, weight and stick within the stocks available locally....but you risk to many compromises. The cost of the steel while
not insignificant, is still actually a smaller part of the overall build. I strongly recommend that you ask your supplier to get some 9mm or 10mm wall thickness sections for you.
If you proceed with thinner sections only to find that your machine is rigidity limited you'll be very annoyed with yourself, and short of scrapping the whole structure and starting again
you cannot improve it. The frame is the DNA of the machine. Make that DNA good and you'll use this machine for years, improving parts of it as budget allows. If the DNA is not good
then those improvements are 'throwing good money after bad'.

Craig

[pippin88]

Brazing is all good in theory until you look up the joint fit up requirements.
Brazing will not tolerate a loose fitup like welding does.

It will be very hard to braze a large frame.

[peteeng]

Hi Ben - The chances of taking a skim off SHS over your near 3m is slim. There are two reasons to use the flat bar 1) to provide enough material to guarantee that the machining will have enough material and 2) to locally stiffen the area where loads transfer thru the structure. 5mm tube suddenly becomes 3mm in spots if the tube is bent or cupped or crowned once you need to machine it flat. Plus you have to get the two sides into agreement so if one longitudinal is up and one is down you could have to deal with 3mm difference. Be generous with the flat bar thickness will save you a lot of angst.... Peter

Hi Pippin braze welding is like tig welding. Flow brazing requires tight fits. You can braze with a mig welder and I think that is what is being discussed which requires fits similar to welding fits. Peter