Well here goes... Aluminium frame router design

[Boyan Silyavski]

Maybe i am being thick but i cant see how flipping the rails will introduce vibration, looking at the conventional method then the spindle is mounted away/lower from the bearing blocks whereas my design is a constant length of zero from the base of the spindle clamp to the bottom of the bearing block. Is it the spindle clamp causing the issue or is the spindle too low in the clamp? How low can the spindle be in its holder?
Cheers

Whatever you do if you dont:

-use long size 20 bearing blocks
-space the bearings that the front plate slides on at least 260mm from outside ends/each long block is 90mm long/
-fix the rails on the front plate back

you will finish with a machine incapable of working Aluminum and maybe not so precise on hard woods.

There not so much to understand, the rails +the bars they rest form ribs that prevent long axis bend. That simple.

Front plate must be 12mm steel backed with steel bars below the rails, or at least or 20mm aluminum backed with aluminum bars. Using 2 spindle brackets would make things even better as the spindle will add stiffness.

Linear rails are cheap, so that will add very small amount to the build. See 200mm travel axis made same way as we speak that reaps/i can not help say it / fully extended , through aluminum without any chatter . here post #109 and how looks finished from 12mm steel plate a bit up at the same page

[CharlieRam]

Hi Silyavski, the rails and bearings are 20mm profiled from chai, they are 75mm long plus 15mm for grease nipple which I haven't added to my sketchup model yet. I originally spaced the bearings 250mm apart from outer dimensions but was told 200mm was fine, this then gave me 150mm depth of cut instead of the original 100mm I planned so I have no issue going back to 250mm spacing if it will help.

The front Z plate is 20mm thick steel with 15mm X 50mm steel spacers glued and bolted to the plate and then the carriages are bolted onto this, the way I see it is the spacers should give me the added rigidity, would it help if they were surface hardened to make them stiffer because that is the only difference I see between having the rails mounted to the Z plate.

I am just finding it hard to understand why flipping the rails will introduce vibration as dean stated and I can't understand how the plate can bend as it has the bearing blocks on each corner. I am planning on having two spindle clamps as you suggested, in fact, it was there on earlier iterations, I must of deleted it!

Jazz, yes it is the workpeice that resonates in my example but the rubber absorbs a lot of it and I am thinking if I tensioned the rubber over the front of the Z plate it would absorb some of the resonance created in the steel Z plate if that is indeed the issue (steel instead of ally?).

Thanks, Charlie

[Boyan Silyavski]

Given the short length of your desired travel it will work ok. Bus as you have drawn it, the bit should touch the finished machine bed. Cause lower it will not go obviously. Is it so?

The case is if say you want it to travel 200mm for example or longer , then the 4 blocks on the Z will go higher than the gantry, hence the tuning fork.

[CharlieRam]

The case is if say you want it to travel 200mm for example or longer , then the 4 blocks on the Z will go higher than the gantry, hence the tuning fork.

So, is that the issue that will cause the vibration, the rails going above the gantry top? If it is then I will have a look at what I can do.
Cheers, Charlie

[JAZZCNC]

I am just finding it hard to understand why flipping the rails will introduce vibration as dean stated and I can't understand how the plate can bend as it has the bearing blocks on each corner.

Charlie it's not rocket science mate. It's all about leverage and vibration. Quality of finish mostly comes from trying to get the lowest amount of vibration at the tool. So we use the shortest tool we can with the least amount of stick out from the spindle to shorten the lever.

Your design is a very long constant length lever that never changes no matter what depth we cut.? Your lever length is distance the spindle extends from the spindle mount plus the tool extension. Lets say total extension of 150mm.

Now if you swap the rails around your lever is variable length lever and only ever reachs 150mm when cutting really thin material and at full extension.
Every where in between and the leverage drops and the stiffness increases so there's less deflection at the tool. At full height our extension is pretty much just the spindle nose and the tool length so order of magnitude better than your design.

May seem like a small thing to you but it makes all the difference when things get hard.!! . . . . Your design is ONLY good for wood anything harder will show it's weakness quickly.
Also with your design then using thick steel for the back plate will add very little to the design other than making the Z axis motor work harder lifting more weight. Your gaining very little strength over aluminium because it's supported at the 4 corners and not extending, if it was extending like a lever then yes it would give more strength.

Regards the size and use of machine then I'll repeat what I've said many many times it's always best to build a machine designed to do the job intended and do it the best it can. JACK of ALL TRADES is always a compromise in some department and this compromise increases with size.
Small machines do it best because they are easier to OVER build.!

[routercnc]

Hi Charlie,

Good to read that you've switched to the conventional Z axis, although I hope you've done it due to understanding the problem rather than peer pressure.

To to lay it out again, using the calculations for the simplified example previously shown:
Conventional Z axis deflection of between 0.19um - 10um depending upon Z axis extension
Unconventional Z axis deflection of 10um - FIXED (i.e. the worse case conventional at all times)

Your unconventional sketch in post #39 looked like it might get around the 'tuning fork' problem but it would not be very practical as it was drawn at full extension and this meant cutting would be up near the gantry! Your only option for practical work would be to drop the spindle down in the mount to hang down out of the clamp, but I think this would not be especially stiff. Your only other option in the design would be to raise the lowest bearing block to allow the Z axis to drop further but then you are back around the loop of an unconventional Z axis as per the above.

[CharlieRam]

The compromise i am making is due to lack of space for the M/C vs. Large enough work area. With this design the total width is about 900mm with the other it was about 1100mm. I havnt even got anywhere to put the machine yet, I have an area of 2m x 5m behind my log cabin which i plan to build a soundproof ish shed with concrete base for the machine, it will then run lengthways leaving me roughly 1m workspace to pass the machine, the other design will only give me about 800mm which is just too small.
TBH i am building this machine to prove i can and to give me something to do. It may well never even nachine ally but then again i might dream of all sorts of wonderful things to make. Who knows? :-)

[CharlieRam]

The machine overall size is 950 wide by 1550 long with a cutting area of 610mm x 1300mm x 150mm
The upper cross brace is also used to mount the stepper but I can add more strutts between the x rail and 50mm frame if thats what you mean?

[ba99297]

Hi
First of all be polite with my English

About your plan
I think that the tube where the rails rest, should be wider (60mm at least ). When you apply the epoxy you will need 20mm for the rail and 15mm-20mm left right of the rail for the meniscus effect ( the tension that has the epoxy at the edges ) . So your tube should be at 50-60 wide minimum)

At the top view I saw 8 diagonal braces. I think you will need only 4 at the edge corners.
I see that you have the y axis ball screw at the middle protected by a plate. Think how you are going to work install and adjust the ballscrew as many time we think “I will think about it later” and finally you find out that you cant even bolt is in place”


Things you must be careful (others told me do so and they are right)

Spot weld in order not to produce too much heat. Even with 80X80X4mm distortion is an option ). After the frame get a first shape you can make bigger welds, no more 3-4 cm at a time. I know everybody want to make big weld with many amperes at the arc welder but this produce too much heat.
Try to be as accurate as possible at every step, even if you finally apply epoxy. Of course don’t make excesses.


It is very similar to mine machine that I design for about a year and start to build 2 months ago


My build starts at page 12 of the thread

http://www.mycncuk.com/threads/6674-...-please/page12

My main specs are
80X80X4mm tube for frame and gantry (gantry at the middle has a piece 120X60X3mm )
80X40X3mm steel tube for the bed.
Working area
750 X 1400 X 180mm

4 Yaskawa 400 watt servos (2 for x axis )
20mm profiled rails

The design of the table was mainly based on syliavski construction.
Also the one who help me most is Dean. I change my mind many times, I get the basic idea for the table from Syliavski, Dean answers all my questions and I think that I make a good start. Of course i made mistakes. I think it is very difficult to avoid mistakes at the first construction. It is very important to have people that inspired you by their ideas, also people who are there to help whenever you have a question.It would be wrong if I didn’t mention the help of Jonathan for his servo size calculator, and Eddy Current for his information and all the people that ansewr to my thread.

Finally I want to suggest you to think how you are going to move that monster. First things are easy but as you add weight it would be difficult to move and turn the frame. Mine now is nearly 250kgr. If I didn’t install a winch it would be a nightmare to move it. You can see the photos to understand what I mean

. http://www.mycncuk.com/threads/6674-...-please/page13


Have a good start

Vagelis

[CharlieRam]

Hi, I did read about the meniscus effect but I am sure I also read that 10mm either side was enough which is why I went for the 80x40, I could lose the 4 inner angle braces and possibly use the spare material to add another cross brace. It is very similar to your design except yours looks a lot stronger, my ballscrew is actually bolted to the front ecocast plate and so 'should' run parallel to the rail front face meaning I only have to adjust up/down, a couple of holes through the box section gantry should enable me to adjust that unless you can get bearing supports that are threaded?
Yes the frame will be heavy but not as heavy as yours, I will need one length of 80x40x4 and 2 lengths of 50x50x3 which I believe will come to just under 120kg, that is without ballscrews etc so should be managable for a couple of people.
I've never welded anything before so I will be practicing first but I have read about spots and little bits at a time so wish me luck on that!
Cheers, Charlie