PDA

View Full Version : BUILD LOG: Comments on my first attempt at a design (if you would)



lukekjackson
28-01-2021, 12:58 AM
Hi,
I'm almost done with what i think is a reasonable design for a benchtop CNC mill. Bear in mind that i say that with no real idea what I'm talking about. I started off with an SX1 manual which i wanted to improve and the project, "grew legs" as they say.

My hope is if I post a few shots of the model and some general info you more experienced folk will be able to give me some tips / highlight the weak spots. I have taken note of the point that is often said that there is no replacement for mass so while this is a benchtop machine it will have to be a fairly solid bench.

To put some words to the pictures-

The column is 2 No SX3 columns which are 100 x 50 cast iron box with a 10mm wall - width over the two 270mm
The Gusset plates at the back are 10mm steel
The Z Risers are 100mm high 25 x 100 steel with a 25mm plate on top for the columns to fix to.

All the ways are linear slides all with 4 carriages
Z 42 x 10 THK, cant remember the model
X Hiwin 20mm
Z Hiwin 25mm

All Axis have Ballscrews
Z 1605 from CNC4you
X & Y 1204 probably from BST Motion

The Base is a cast iron Surface plate with a 25mm thick top plate

the belt drive is a DIY to give 2:1 up gearing and about 2:1 down. driven by a 500w BLDC sewing machine motor. which runs 200-5000 RPM

I hope to be able to cut steel, but have no particular speed in mind.

Solidworks puts the mass at C.142kg

The base is 24" x 18" and the Z is about 600 high.

Any questions or comments welcome

29438

29439

29440

29441

Cheers,

Luke

routercnc
29-01-2021, 05:59 PM
Looks like a good start. Some comments from me:
500 Watt motor seems a bit underpowered if you want to cut steel regularly. It is BLDC which should help so is it a special high torque version ?

The plates holding the pulley assembly look cantilevered off the motor body. Can they join back to the black housing to stop them vibrating?

Main black housing is quite narrow and the footprint ends up some way inboard of the Z rails. Can this be widened at the rear or diagonal bracing added?

Think about how you will cover the rails with bellows or way covers.

Think about if you will add flood coolant and provision for a tray around the Tslot table.

Is the space above the spindle for a future power draw bar? If not make sure there is space to add one.

Make sure there is provision to tram the spindle and column relative to the table. Either a fancy method or bolted joint with shims.

Otherwise looking good.

routercnc
29-01-2021, 06:02 PM
Double post sorry

magicniner
29-01-2021, 06:10 PM
Double the vertical size of your Z axis carriage, at least.

Muzzer
29-01-2021, 07:06 PM
Yes, the carriage height in Z (top to bottom of the moving body) is very short and quite a bit less than the overhang of the spindle. Also the depth in Y from the spindle to the column is fairly limited. Once you have a vise in there, you will lose some of that, reducing it even further. The Sieg cast columns are standing on spindly pillars - solid blocks would be sensible.

Looks like a plan though - get cracking and show us how you get on! You are bound to evolve the design as you go so the sooner you start the better.

JAZZCNC
29-01-2021, 09:44 PM
Everything the others mention plus a few more.
#1
Widen the Y-axis rail spacing, you have plenty of space so use it. Also, this will allow easier access to the bearings bolts, the way you show it now looks like you the X-axis bearings will block access to the Y-axis bearing bolts.?

#2
The triangle plates at the rear I would move to the outer edges and fasten to the side of the columns, this will give a wider footprint, I'd also think about adding some lateral bracing to add sideways support. Like muzzer mentioned the column risers look a bit flimsy so maybe beef these up.

#3
It's been said but worth mentioning again because it's a Biggy. The Z-axis plate needs to be double the height and the spindle housing needs to be wider.
Also, the spindle motor plate needs supporting or the vibrations will kill the spindle bearings and vibrate on the machine general.

Good luck.

hanermo2
29-01-2021, 10:17 PM
The z axis ratios are way wrong -- as others have said.

You need about 1.5 x width/depth vs height for best results.
Aciera comes to mind.
Deckel.
Schaublin.
About 2x for most commercial machines.

So if you have 600 height you should have 300-400 mm width and depth for best results, and not less than about 250-300 in each dimension.
Itīs dirt cheap , relatively.
All modern machines use similar methods with large sections and relatively thin walls around 10 mm.

lukekjackson
30-01-2021, 01:18 PM
Thanks to everyone for the responses. You've mentioned a couple of things that i'd considered but definitely brought some new ideas up as well.

RouterCNC,


The whole head is a weak point i think, The project started off as upgrading a Clarke CMD10 and the only bit that remains in the head and spindle as part of tis design. I think ultimately it will be changed later on for a bt30 spindle.

29442

This is the head from the CMD it has a 30mm travel on the Z for drilling and the spindle moved relative to the head, that's the reason the pulley drive is canter levered that way. I left that function as it was there but on reflection it wont really be that useful on CNC so i can drop it to its lowest height an fix the back of the pully base plate to the head.

The 500W motor was to be an upgrade from the 150W motor on the clarke mill which seemed a reasonable size for the mill, Its a JACK sewing machine servo motor but think it has gone the other way now. I think at this point id prefer to have the motor underpowered to the mill frame than the other way round but i may upgrade the motor later depending on the spindle



Main black housing is quite narrow and the footprint ends up some way inboard of the Z rails. Can this be widened at the rear or diagonal bracing added? - I'll add diagonal bracing.

Think about how you will cover the rails with bellows or way covers. - I do intend to include this but haven't modelled it yet,

Think about if you will add flood coolant and provision for a tray around the Tslot table. - Yep the plan is for this to be in an enclosure the floor of which will fall to a central drain.

Is the space above the spindle for a future power draw bar? If not make sure there is space to add one. - Will do

Make sure there is provision to tram the spindle and column relative to the table. Either a fancy method or bolted joint with shims. - [B]This ive been thinking about, i can work in provision for adjustment even if it is, like you say, shimming but I cant work out how to measure it[. I can tram the head with a dial checkng front, back and side to side in the normal way but what worried me is ensuring the Z rails are vertical. If not it may tram ok but at the head moved in the Z it will move relative to the X and Y. ANy tips on that would be great/B]

lukekjackson
30-01-2021, 01:24 PM
magicniner.

Thanks, when i first modelled it i had the wrong linear rail blocks and the height of the z carriage came out at 175mm, when i actually put it together ang butted the blocks up it turned out it was 100, so I re-modelled it to suit.

It did strike me that is looked wrong then but I've already made the dam thing.

I think that will be the first job of the mill, to make a new Z carriage mounting plate. Ill change it on the model to make sure nothing else will clash later on

lukekjackson
30-01-2021, 01:44 PM
Muzzer

Cheers for you comments. The one about the pillars surprises me i thought 25 x 100 solid steel bar would be over the top if anything. My logic was that the columns only contact at those points and are relieved in the middle of the base, so with the 25mm thick flat bar fixing the top of the risers, and the plate at the bottom i thought it would be as strong as having them on straight to the base plate.

carriage height in Z (top to bottom of the moving body) is very short and quite a bit less than the overhang of the spindle. Yes I will increase this to 150 - 200mm from the 100 it is now

Also the depth in Y from the spindle to the column is fairly limited. Once you have a vise in there, you will lose some of that, reducing it even further. - I tried to space it so the axis of spindle was relatively central to the limits of the Y travel, see below. Would it be more useful further out. Lack of experience on my part to know this

29443
Spindle axis to table edge - 23.5mm

29444
Spindle axis to table edge - 10.42mm

lukekjackson
30-01-2021, 01:59 PM
JAZZCNC, thanks

Widen the Y-axis rail spacing, you have plenty of space so use it. Also, this will allow easier access to the bearings bolts, the way you show it now looks like you the X-axis bearings will block access to the Y-axis bearing bolts.? - One iteration actually has these wider, not sure why i closed them up but i will move them back out,. Nice easy win there

This brings me to one thing I've been struggling with, the X axis has a lot of overhang. Using 25mm HIWINS with the block spaced 36mm apart does anyone know what the max overhang should be? Obviously the longer i make the table to more travel but there has to be a point where you lose rigidity

29445

#2
The triangle plates at the rear I would move to the outer edges and fasten to the side of the columns, this will give a wider footprint, I'd also think about adding some lateral bracing to add sideways support. - Again another easy win. the ridiculous reason they are there is that the column came with holes tapped for the Z rack so i used these. it will actually be far easiest fixing them to the sides as i don't have to drill through the depth off the gusset plate,. Should definitely have been doing it your way in the beginning

Like muzzer mentioned the column risers look a bit flimsy so maybe beef these up. - Still surprised but you all seem to agree. this may be one to do when i have the mill working so i can skim the sop of some stock to flatten it out. Either that or i could fit letteral section in between the risers to strengthen the riser design

#3
It's been said but worth mentioning again because it's a Biggy. The Z-axis plate needs to be double the height and the spindle housing needs to be wider.
Also, the spindle motor plate needs supporting or the vibrations will kill the spindle bearings and vibrate on the machine general. - Ill pick these up on the next revision and repost a snapshot of it

Good luck.- Cheers, Ill need it

John11668
30-01-2021, 02:13 PM
I wonder why you felt you needed such column height relative to your X and Y dimensions . What height of components do you envisage .

One of the most difficult machining tasks I remember from my early days in industry, was a tall titanium columnar forging which needed double curvature machining onto its top face. It cost a fortune in fixtures to fix and support it and the machining had to be extremely slow to keep cutting forces within bounds. It was done on the table of a triple headed Bridgeport copy mill ( pre cnc days) So machining time for the op was 300 hrs although we could produce 3 in that time, and we needed a pair for each airframe. Prima Donnae in the drawing office would not consider a design change which could have cut the time to a tenth.

The point here is that columnar components are seldom machined standing upright on the table so large throat clearance is seldom needed .
It detracts from rigidity in the column and also in the component , so you get a double dose of flex in response to the cutting forces .

What is the maximum height of your proposed component? Work from that taking into account the Z positions for clearance and depths of machining operations , spindle retraction etc . Tall components will need to be braced from the table so add to the need for table width and extra rigidity.

Looking at your design I see an X / Y table sitting on a flat base . You could save yourself some time by buying one of those and adding the ball screws and motors which was the route I chose to take. OK so mine may not be the prettiest in the world but I wanted to get quickly to the eletronic/ software stage and start to play.

Hanermo mentioned Deckel Aciera and Shaublin , all good examples of well designed column mills from which design lessons can be learned .
I used my Deckel FP1 as a base , but as they say on the BBC , other makes are available . You will not find anything more rigid upon which to base your build .

My project is here http://www.mycncuk.com/threads/13746-Compound-table-conversion-of-an-old-Deckel-mill?p=117802#post117802

lukekjackson
30-01-2021, 02:14 PM
hanermo2, Cheers

The z axis ratios are way wrong -- as others have said.

So if you have 600 height you should have 300-400 mm width and depth for best results, and not less than about 250-300 in each dimension. - Column Width at present is 270, Y to the furthest point of the gusset from the face of the column is 190mm
Itīs dirt cheap , relatively.
All modern machines use similar methods with large sections and relatively thin walls around 10 mm.

I think I'm a but stuck with this as i already have the surface plate. I thought as the column was off an SX3 and i was using two size to side it would be far stiffer in the Y axis then the gusset plates would help the Y. I will change the gussets as advised by Jazz and may put some in the x axis but I've run out of depth in Y to stiffen it up any more, i don't want to move the column forward as i will loose Y travel.

What would the consensus be, move column forward and loose travel but stiffen the Y or not?

lukekjackson
30-01-2021, 03:11 PM
John11668.

I wonder why you felt you needed such column height relative to your X and Y dimensions . What height of components do you envisage .

In short i didn't really decide on relative dims, i wanted to upgrade the Z on my CMD10 so bought a a X3 column which has far better proportions than the stock CMD. While tapping for the linear rail fixings messed up a couple of holes and snapped a tap in 1 so i bought another one. Then as i had 2, I thought if I fix them side by side its even more rigid. once i modelled this up it looked ridiculous in comparison the the X/Y table (already replaced from the original CMD10 one). so whats the the obvious next step, start on a new X/Y of course. It is a slight case of scope creep from the original intent


What is the maximum height of your proposed component? Work from that taking into account the Z positions for clearance and depths of machining operations , spindle retraction etc . Tall components will need to be braced from the table so add to the need for table width and extra rigidity. - I have been thinking about this and you are correct i cant see myself ever needing to mill the top of something 400mm + high in a serious way. What i could imagine though would be drilling or flattening an edge. I'm effectively thinking i may mill up to a couple hundred mill high and only drill above that

Would it be worth losing the risers, i just didn't like the idea of having unusable rail under the lowest point the head can reach. and if the highest point of the Z axis is rarely gong to be used or only used for drilling does it matter if when using that section it is not as ridged as ideal?

Looking at your design I see an X / Y table sitting on a flat base . You could save yourself some time by buying one of those and adding the ball screws and motors which was the route I chose to take. OK so mine may not be the prettiest in the world but I wanted to get quickly to the eletronic/ software stage and start to play. - that was my initial plan and i did buy a larger X/Y table, i Want to add linear rails and ball screws though and when i though about what i would have to do to the X/Y, and how i would do it only having access to the CMD10 to do the work it seemed like it would be more effective to use a surface plate and ground flat stock to build one. It would save me a lot of machining squaring things up in theory. I have no doubt i will live to regret that.

Hanermo mentioned Deckel Aciera and Shaublin , all good examples of well designed column mills from which design lessons can be learned .
I used my Deckel FP1 as a base , but as they say on the BBC , other makes are available . You will not find anything more rigid upon which to base your build - Just had a look at the FP1, that does look look like a solid lump. To give me some frame of reference what depth and speed of cut would this sort of machine regularly make in steel. the Belt drive I've included will take the spindle speed up to 10K RPM so i'm of hoping to use smaller cutters at higher speeds to make up for the lack or rigidity of not having a big cast base

John11668
30-01-2021, 04:28 PM
These mills are toolroom mills http://www.lathes.co.uk/deckel/page5.html
Shows the sort of work they normally are bought for . Lots of clones including Alexander in UK https://www.ebay.co.uk/itm/164561029251?ViewItem=&item=164561029251
and optional high speed head will go up to 6K rpm but lots of scope for bigger cutters /

These are pricey but lots of other cheaper options Harrison, Viceroy, Elliot to do similar with

lukekjackson
10-02-2021, 12:20 AM
I'm back with my updated "design".

I think I've taken most of the advice on board, the glaring omission in the "spindly" legs. My problem with this is the ability to flatten a solid block of steel that size. I've added another leg, actually due to the fixing points clashing with the webs under the Surface plate, but it cant hurt rigidity. Once its up and running i may revisit this.

General Overview
29513

Increased Z Carriage
Head Strengthened Laterally
Pully Drive Fixed down to the head at the back
29514

Column Gusset plates moved
29515

Y Axis Linear bearings widened
29516

I've also lost the T=slotted table and decided to go with a fixture plate type setup tapped for 6mm bolts and drilled and reamed for 6mm dowels for alignment. The table has got longer though which makes the overhang a bit of a worry.

29517

My reading of the HIWIN Data sheet for the HGH25 say dynamic moment My is 240Nm equates to c. 24Kg at a meter for one block, my total distance from block to furthest point is c.0.5m meaning 48kg is required to hit the max, there are 2 rails, so 96Kg and that doesn't account for the blocks being in pairs per rail, which you would assume would have a fairly large impact torque applied.

So if i understand the Data i could put over 100kg on the table at full extension and still be in the limit. If that is the case i have no worries but it just looks wrong. Could someone check my logic.

29518

The last lot of comments were a big help so any more would be appreciated.

I'm also looking at a frame and enclosure so will be back with more at some point

Cheers,

Luke

John11668
10-02-2021, 12:41 AM
Starting to look almost like a Deckel :peaceful:
So cant be bad !