Need Advice on this Table Top machine design

[swill008]

Current Project [Mockup] SnapShot


Im looking for advice on this build. Looking at the structure to see if it is rigid enough.
I do plan on having the lower assy as an torsion box [Frame it in really well] using 80/20 products.

Any advice or cautions/recommendations for this build?

I was inspired from Nero7 on YouTube and decided to take the plunge [Hobby].
I do plan on using the electronics/motor package and scale up the project to a 4x8 when feasible for me to do so.

The frame is a 29x36 using a Bosh router for a spindle. Using linear rails and a KR33 linear actuator for the z axis.
I'm still in the design stages for this before I start ordering parts. So far, I haven't chosen the parts yet [sizes] so
a lot of the structure is not complete in the model. The table top will be MDF and Ill frame in the base with 1x2 so I
can mount it.. [well that's my plan...]

Please feel free to critique my build and design. My intended use is hardwoods and aluminum milling.
I am trying to make this little guy rigid as possible [within reason], and use the design to scale it up.


Thanks!

SolidWorks 2015 Go Pack, with all the models and Assy [First incomplete Draft]

Similar Threads:

• My first DIY CNC machine design please need advice
• New Machine Build - Need design advice
• 1x1.5m table, new machine design
• New Machine Design book from Alexander Slocum (of Precision Machine Design)
• XY table design basics, linear ways to table size ratio?

[swill008]

Looks like I'm getting a lot of hits, but no responses!

I would really like your opinions, good - bad - ugly.... what ever


Any questions or if you want me to export my files to something more usable for you to look at - Just let me know!

I plan on using on the Y axis linear rails, NSK that are @29.75 inches long and 15[larger?]mm wide with 2 trucks per rail.
X axis linear rails NSK @23.75 inches long and 15mm wide with 2 trucks per rail

X axis ballscrew THK, @ 26 inches of thread and is @ 3/4 inch in diameter.
Y axis ballscrew Dual THK or NSK, @ 32 ish inches of thread and is @ 1/2 inch in diameter.


Z axis uses an all-in-one THK KR33 actuator with integrated ballscrew that has @4 inches of travel.

I know i would not need Nema 34 motors, but i rather go larger and move this to a larger build if feasible in the future.

[UA_Iron]

Can you upload the model in a parasolid model? I'm running Solidworks 2014 and won't be able to open it.

What materials will you be cutting and what kind of tolerances do you expect from the machine? I suspect wood, mdf, possibly aluminum...

Extruded aluminum has a noticeable amount of twist throughout the length of the extrusion. Misumi offers ground extrusion if you need to hold tighter tolerances. Supposedly, Bosch has a very advanced extrusion method that allows them to use them for their stages without any post grinding... though I have not seen these available to consumers like the 80/20 and misumi extrusions.

Your current design has some areas that *could* be optimized (really, this is requirement dependent). Structural steel beams will offer the most rigidity (for the DIY CNC application). You want to populate material in the cross-section opposing the forces. The plates on either side on what you defined as the X-axis may be subject to bending, but they are rather thick. A structural tube member would work better and offer more rigidity, but would be much more work to implement and integrate. Loading transferred in the X-axis direction is going to cause a moment about the y-axis rails and cause some kind of deflection; really another set of linear y-axis rails placed below the y-axis ballscrews and connected to the x-axis supports would counter this moment very well. But, like I said, this is requirement dependent. If we kept going down this optimization path you might end up with a HAAS CNC.

[ger21]

I don't have Solidworks, and your pics are a bit small to really see anything.
One thing I do see, is that the spindle mount to Z axis connection might be suspect, but it's hard to see.

[wizard]

Looks like I'm getting a lot of hits, but no responses!

I would really like your opinions, good - bad - ugly.... what ever

Give a guy a chance! Some of us aren't here everyday😃😉😉😉

The first thing that comes to mind here is that it might not make sense to design such a machine to scale up in the future. You would end up replacing just about everything at which point it might make sense to optimize this design for today. This is a tough call because if the machine was a bit larger, say 2x4 feet, it might make sense to put a bit of money into a 4' long gantry. That would be sub optimal on a 2x4 foot machine but if you knew you wanted to make a larger one it would make atleast one part of the machine transferable.

On another note don't assume that you will want to get rid of the desktop machine once you have a larger machine. Small machines can be built to be very rigid at moderate costs. The other thing is that you might want to consider an all metal design if you intend to do much aluminum machining. The reason here is the use of coolant/lubrication while machining aluminum. Obviously I don't know your intentions with respect to aluminum but if you are looking at a significant percentage of your usage I would avoid all wood products in the build.

Another benefit of keeping a small machine around is that it is easier to totally enclose the machine. This can have huge advantages in controlling dust and airborne fluids if you do end up machine a lot of aluminum. You can easily create a recirculating lube system if the machine is fully enclosed. Also an enclosure helps with noise control. I just wanted to point this out as it can justify keeping a small machine around.

Any questions or if you want me to export my files to something more usable for you to look at - Just let me know!

I plan on using on the Y axis linear rails, NSK that are @29.75 inches long and 15[larger?]mm wide with 2 trucks per rail.
X axis linear rails NSK @23.75 inches long and 15mm wide with 2 trucks per rail

X axis ballscrew THK, @ 26 inches of thread and is @ 3/4 inch in diameter.
Y axis ballscrew Dual THK or NSK, @ 32 ish inches of thread and is @ 1/2 inch in diameter.


Z axis uses an all-in-one THK KR33 actuator with integrated ballscrew that has @4 inches of travel.

I'm not familiar with that actuator and can't look it up at the moment. So all I can say is consider carefully its rigidity. Many linear actuators on the market these days are not really designed for machining loads. Rather they are often employed as pick and place solutions.

I know i would not need Nema 34 motors, but i rather go larger and move this to a larger build if feasible in the future.

Again I'm not sure in this case it would be a good idea to design this machine with a future larger machine in mind. You will end up tossing just about everything. Further you may want to keep this machine around for the larger machine build.

As for the materials, a machine this size is fairly easy to do in extruded aluminum. As noted above I'd avoid wood if you have a significant amount of aluminum planned for the machines future. I'd also seriously consider steel as it is much cheaper though maybe not glaringly so on a machine this size. The thing with steel is that you can weld up a really stiff and robust frame that after a bit of stress reliving out to hold shape for a very very long time. The gotcha is the need to do post welding machining though there are ways around that need. The other advantage to a steel frame is the ability to fill that frame with concretes to add mass.

The big problem with steel is getting the machining work done at a reasonable expense or getting access to equipment to DIY. In that regard do consider local Maker Spaces, TechShops and the like. Such facilities, assuming they have the equipment, can be worth the price of admission and would allow you to do steel construction in traditional ways. If you don't have such access the third alternative is to use some of the processes that have been developed here and other places over time. One such approach is epoxy leveling and grouting, to produce precision surfaces and to achieve precise machine alignment.

As to aluminum T-slot extrusions I have mixed feelings when it comes to the use of these extrusions in a machine build. As noted they are expensive and people often dismiss the need for gusset plates and the like which adds to the expense. On the other hand you can build a perfectly good machine in this size with extrusions. Just pay attention to how everything is assembled making sure that each and every joint is properly torqued and is properly supported to prevent twisting. I would want to see a flat plate gusset at every joint or at a minimal a corner bracket. To some that may sound like over kill but I've seen this stuff twist and otherwise go out of alignment at the joints.

One last thing. A router might not be your best bet for machining aluminum. The spindle isn't a robust as it could be but my concern is the spindle ingesting metal and shorting out.

[swill008]

I plan on using the KR33B [2 Blocks], my model only has the one showing.

Click here for the Manufacture's KR33 PDF

UA_Iron,

I have attached the parasolid file of my mockup model. I will be using this up to aluminum, the MDF top
will be replaced eventually with an aluminum top eventually so that I may run coolant [Enclosed]

I would prefer precision over speed, Mostly I want to be able to mill aluminum parts accurately and
as feasibly able to do so, but as cost effective as possible. So I want my cake and eat it too! lol

I plan on using Ground Ballscrews, +-.001 is what I would hope for.. in the accuracy of the machine.

You recommend I create a C-Clap style with another rail directly below the y axis rail? just trying to wrap my
head on it. I could see that - I guess I was thinking the ballscrew would take some of that mentioned
force.


Ger21,


I attached another render of the z axis and uploaded these to photo-bucket for higher res images.
Let me know what you think or have a look at the parasolid file attached. [Please keep in mid that
its a partial mockup - no fastener holes ect.. and ill be using a 2 block KR33 Actuator, not shown.
[Haven't found the dimensions for my exact travel]


Wizard,

I know you live on these forums looking for guys like me to pick your brain!! hah [Sarcasm, btw ]

I agree almost every part will be replaced. I needed to be more clear, I was hoping that the design
would scale to 4x8, but the more I read the more I'm thinking towards UA_Iron - using structural
steel for that purpose.

I would just assume based on my current scale that nema 23's will work well on this design, so the
nema 34 should be overkill, but would transfer over to another [larger] machine. I don't want to
purchase all new electronics.

[Probably would as you mentioned, keeping this machine anyway! - The wife would love that!]

As mentioned previously above, I want enclosed cooling [recirculating] and the MDF top replaced
with metal. I would like to do as much as I can without having to machine surfaces, but I can also
be a realist and know most of my design calls for a CNC to cut it out. I'm sure I can find a willing
partner here somewhere!

Maybe obvious to you, but I assume it would be cheaper for me to stay smaller - than to scale
this build up to say 2x4 or 3x4 cutting area?

I guess I'm looking for optimum design space for common components, but It seems it just a lot
of any dimension parts just floating on ebay ect. I tried to constrain my build to parts I could find.

I do plan on keeping the 80/20 products well braced and gusseted, also I would upgrade the
router when it was feasible depending how much aluminum i actually did mill.


Thank you! All of you for your input and comments.

I have very thick skin and don't hesitate to call me out! Or tell me it - JUST wont work... lol

[ger21]

I would prefer precision over speed

Everyone says this when they are building their first machine. They really have little to do with each other.
A fast machine is not necessarily any less precise than a slow one. It may have a theoretical higher resolution, but that doesn't mean it will be more precise. Precision has a lot to do with the quality of the components, and the quality of fabrication and construction.
What you really need for precision is repeatability, and it's not that difficult to get good repeatability.

Also be aware that you can't always just cut slowly. Different materials require different feedrates, and if you can't achieve those feedrates, then your cut quality and tool life can be greatly diminished.

I was hoping that the design would scale to 4x8...........

No design of that size will really scale up to a 4x8, for a variety of reasons.

Really, about the only parts that would scale up would be the gantry sides, but even those would probably need to change. The gantry beam on small machine would likely not be nearly rigid enough when you double it's length. A larger, more substantial gantry beam would likely require changes to or entirely new gantry sides.

You're ballscrews are not going to scale up at their current diameters. New, larger ballscrews may require significant changes to accommodate them.
Motors that are optimized for your current ballscrews are unlikely to work well with much longer, larger diameter ballscrews. What pitch are the screws you plan on using? For a 4x8 machine, you'd probably want to use a 25mm pitch.

As parts get longer, they get more flexible. I would guess that the entire frame would need to be much heavier duty, especially if you want a 4x8 machine for cutting aluminum.

If you really wanted a machine that could scale up to a larger machine at a later date, then it would not make a lot of sense to even bother with the smaller machine. Most of your cost is going to be in linear bearings and ballscrews. The additional materials for the larger frame will have a minimal cost impact. By going with the large machine first, you'd save a considerable amount of money by not having to re-buy all the linear motion components.


Also, if cutting aluminum is your main goal, I'd look for a heavier duty spindle than a Bosch.

Post more closeup pics of the different components, from different angles. 1 side pic doesn't show nearly enough.
And I can't open a parasolid, either. And don't really have time to open and look through a model.

[wizard]

I plan on using the KR33B [2 Blocks], my model only has the one showing.

Click here for the Manufacture's KR33 PDF

UA_Iron,

I have attached the parasolid file of my mockup model. I will be using this up to aluminum, the MDF top
will be replaced eventually with an aluminum top eventually so that I may run coolant [Enclosed]

I would prefer precision over speed, Mostly I want to be able to mill aluminum parts accurately and
as feasibly able to do so, but as cost effective as possible. So I want my cake and eat it too! lol

You seem to be refocusing on machining aluminum as the primary usage for this machine. What you will be machining parts wise has an impact on the machines suitability. Doing such parts with precision, in this case I'm suggesting sub one thou then you will need to put a lot more effort into the quality of the build. This means machined surfaces even if you go to T slotted extrusions. I'd have a hard time recommending extrusions if you truly want precision that is maintainable over time.

I plan on using Ground Ballscrews, +-.001 is what I would hope for.. in the accuracy of the machine.

Ball screws are only part of the solution, if your linear rails aren't mounted on flat surfaces you will loose any potential for accuracy. Further I'm not all that thrilled with the idea of getting significant precision out of a moving gantry machine. Now much of these comments must be hedged with the idea that it depends upon what your expectations are.

You recommend I create a C-Clap style with another rail directly below the y axis rail? just trying to wrap my
head on it. I could see that - I guess I was thinking the ballscrew would take some of that mentioned
force.

I'm not sure what is up with that comment either but ball screws are for pushing and pulling stuff.

Ger21,


I attached another render of the z axis and uploaded these to photo-bucket for higher res images.
Let me know what you think or have a look at the parasolid file attached. [Please keep in mid that
its a partial mockup - no fastener holes ect.. and ill be using a 2 block KR33 Actuator, not shown.
[Haven't found the dimensions for my exact travel]


Wizard,

I know you live on these forums looking for guys like me to pick your brain!! hah [Sarcasm, btw ]

Often I'm on break this the unusual times for my responses. In a weird way this forum take my mind off the troubles at work.

I agree almost every part will be replaced. I needed to be more clear, I was hoping that the design
would scale to 4x8, but the more I read the more I'm thinking towards UA_Iron - using structural
steel for that purpose.

Scale - I don't think it is that simple.

I would just assume based on my current scale that nema 23's will work well on this design, so the
nema 34 should be overkill, but would transfer over to another [larger] machine. I don't want to
purchase all new electronics.

[Probably would as you mentioned, keeping this machine anyway! - The wife would love that!]

You never know she may end up using that machine herself.

As mentioned previously above, I want enclosed cooling [recirculating] and the MDF top replaced
with metal. I would like to do as much as I can without having to machine surfaces, but I can also
be a realist and know most of my design calls for a CNC to cut it out. I'm sure I can find a willing
partner here somewhere!

Read the threads from guys that have used various epoxy based solutions to generate flat surfaces. There is a product call Moglice that is worth looking into. The point is you can arrive at a usable machine with alternative methods.

Maybe obvious to you, but I assume it would be cheaper for me to stay smaller - than to scale
this build up to say 2x4 or 3x4 cutting area?

Generally yes smaller is cheaper. However the optimal set of components for a given investment might not imply the machine size you want or need. I've yet to see an indication of what you expect to machine the machine.

I guess I'm looking for optimum design space for common components, but It seems it just a lot
of any dimension parts just floating on ebay ect. I tried to constrain my build to parts I could find.

I do plan on keeping the 80/20 products well braced and gusseted, also I would upgrade the
router when it was feasible depending how much aluminum i actually did mill.


Thank you! All of you for your input and comments.

I have very thick skin and don't hesitate to call me out! Or tell me it - JUST wont work... lol

[swill008]

Hey!

Sorry I don't have much time at the moment but I wanted to post these images. Ill get back to you guys in a bit.
Thank you all very much for all the help and making me think outside my "box"!

I want the ability to mill aluminum without any issues. However I will primary use this to mill hard woods and designs,
so I want to build "up to" aluminum milling. I hope I'm making sense!

[ger21]

At minimum, you want to tie the two gantry extrusions together. Ideally the gantry should be one large beam.
Read this thread: http://www.cnczone.com/forums/diy-cn...cad-posts.html

I think you'll need some type of plate between your 15mm rails and the extrusion, because the rails are so narrow. You'll also need machined extrusions (as was mentioned) to get near your tolerance of .001".

I don't have any experience with them, but those spindle clamps just don't look that rigid to me. I'd prefer something with bracing above the clamp to stiffen it up. MAke sure the bearings are at the top and bottom of the back plate to maximize rigidity.

[UA_Iron]

This is a good thread right here:
http://www.cnczone.com/forums/diy-cn...yy-budget.html

You're planning on using ground ballscrews - are you sourcing them from ebay or buying them retail? These won't be cheap - you could probably get away with a rolled ballscrew type design and build in the accuracy component into your control system using encoder feedback. Even a ground C5 grade ballscrew is not going to give you ±.001" over the length of the table without some kind of encoder feedback. Like Ger21 said, you will be super repeatable with a ballscrew/stepper setup, but those accuracy components will be tough to hit over the distance.

All this kind of steers the conversation to your budget for this machine...

There's a reason why 99% of commercial mills have a moving table over a gantry - it is an inherently stiffer design, but not nearly as space efficient. This is more ideal for cutting metals. Not saying you can't do it with the gantry, but you have to consider that it may be tougher to do over another setup.

[Azalin]

Hi,

This machine look very cool but please allow me to share my 2 cents. Like mentioned before, you need to tie the extrusions on the gantry together. 5mm thick alu plate will do the job. Or rotate one of the extrusions vertically.

The linear rails/blocks on the top are totally open to dust/chips. You'll cut mostly wooden material and maybe MDF. MDF dust is damn thin and I hated cleaning the linear parts after every time I cut MDF. I know it is almost impossible to design an enclosed dust free system but I think, at least you can move the rails to the side. That'll be better.

Best,
Suat

[swill008]

At minimum, you want to tie the two gantry extrusions together. Ideally the gantry should be one large beam.
Read this thread: http://www.cnczone.com/forums/diy-cn...cad-posts.html

I think you'll need some type of plate between your 15mm rails and the extrusion, because the rails are so narrow. You'll also need machined extrusions (as was mentioned) to get near your tolerance of .001".

Agreed ger21,

I have placed .50in back plate and attached plates for the railing. Do you think this would tie things together properly?
My thinking for such a large chunk of aluminum is to have all parts cuts from the same stock, would be cheaper in the long run?
I priced the Bosh milled extrusions, [For 2x4 equivalent] its about 100$ for 1000mm. A bit pricey, so I may have to sacrifice accuracy.
I Cant see spending that much for extrusions, over 400$ for just my rail extrusions, not to mention I would then have to buy all Bosh extrusions so the hardware will line up.

I'm starting to ponder structural steel and epoxy leveling techniques, then go straight to 4x8, but I'm going to try this perhaps for just wood and small hobby cutting If I stick with 80/20 products if my tolerances are crap!

You're planning on using ground ballscrews - are you sourcing them from ebay or buying them retail?
All this kind of steers the conversation to your budget for this machine...

UA_Iron,

I plan on sourcing from eBay, I want ground ball screws, if I can find them for a reasonable price. As far as budget I'm going to build this over a year or so, I have patience so I have no need to rush.
Budget I'm hoping for 1.5 / 2k max. I will also be doing a Bridgeport series 2 CNC for my next project (just need to find one)

The linear rails/blocks on the top are totally open to dust/chips. You'll cut mostly wooden material and maybe MDF. MDF dust is damn thin and I hated cleaning the linear parts after every time I cut MDF. I know it is almost impossible to design an enclosed dust free system but I think, at least you can move the rails to the side. That'll be better.

Best,
Suat

Azalin,

I changed my gantry design to stiffen it up, what do you think? I plan on designing some sort of dust shield to protect my rails. I hope to be a sliding dust shield.
I would prefer my rails on top, but I may try them on the side. Any other cons you can think of with mine on top other than the side?

[wizard]

Hey!

Sorry I don't have much time at the moment but I wanted to post these images. Ill get back to you guys in a bit.
Thank you all very much for all the help and making me think outside my "box"!

I want the ability to mill aluminum without any issues. However I will primary use this to mill hard woods and designs,
so I want to build "up to" aluminum milling. I hope I'm making sense!

You have some nice pictures there. Do realize though that your gantry arraignment might not fully cover machining of the table. This is due to the Y axis saddle being caught between the two uprights. Generally you want to be able to over shoot the sides of the table to flatten waste boards, machine fixtures and the like.

It is hard to tell from your renderings is the spindle will actually cover the entire table top. I'd look into this if full use of the table top is important to you. There are advantages to gantries that allow the router to travel to the outside of gantry supports. It can make tool changes much easier especially if you are short on Z travel.

[wizard]

Agreed ger21,

I have placed .50in back plate and attached plates for the railing. Do you think this would tie things together properly?
My thinking for such a large chunk of aluminum is to have all parts cuts from the same stock, would be cheaper in the long run?

Paying a job shop can become very expensive often exceeding the cost of the materials. Often it worth every penny you pay though. You really have to balance what is achievable by you in your shop against the cost of a job shop doing the work for you.

I priced the Bosh milled extrusions, [For 2x4 equivalent] its about 100$ for 1000mm. A bit pricey, so I may have to sacrifice accuracy.
I Cant see spending that much for extrusions, over 400$ for just my rail extrusions, not to mention I would then have to buy all Bosh extrusions so the hardware will line up.

This is why I suggest people look for alternatives to T-slot extrusions, that and many of those extrusions aren't as strong as people think. If you are looking for the precision that implies machining the rail surfaces you might as well consider steel. In both cases machining is a requirement and the cost to machine will be similar. An outside shop might be able to drill the holes for the linear rails for you on a CNC also. That would cost extra but can save on very tedious work.

I'm starting to ponder structural steel and epoxy leveling techniques, then go straight to 4x8, but I'm going to try this perhaps for just wood and small hobby cutting If I stick with 80/20 products if my tolerances are crap!

If you want a precise machine then build one, if you want one for wood working then build one for that. A well built table top machine can be very useful and if you want very precise. It is really up to you and what you want to put into it.

Note also,that if you want to machine aluminum the stiffer machines desirability goes beyond precision, a stiffer machine means better surface finish.

UA_Iron,

I plan on sourcing from eBay, I want ground ball screws, if I can find them for a reasonable price. As far as budget I'm going to build this over a year or so, I have patience so I have no need to rush.
Budget I'm hoping for 1.5 / 2k max. I will also be doing a Bridgeport series 2 CNC for my next project (just need to find one)

2K for the entire machine with ground ball screws? if that is what you mean then all I have to say is good luck. Further I wouldn't even bother with aluminum if you are going to the expense of ground ball screws.

Azalin,

I changed my gantry design to stiffen it up, what do you think? I plan on designing some sort of dust shield to protect my rails. I hope to be a sliding dust shield.
I would prefer my rails on top, but I may try them on the side. Any other cons you can think of with mine on top other than the side?