My CNC Router Build Adventure

[mactec54]

I forgot to ask, is there much difference between 6061 and 5083 aluminium for my application? 5083 is currently a little cheaper and more readily available than 6061 due supply shortages.

5083 is not as good for machining and has a higher strength compared to 6061

[mactec54]

Oh wow! I have to admit that some days it feels like I am way out of my depth with this project. Today is one of those days. I am currently on a very steep learning curve and it's going to take a while to digest all this information. I'm very grateful for all of your help and shared knowledge, please keep it coming. At some point, soon, I am going to have to commit and start building this machine. I learn better with practical experience than from theory.

Thank you for the ebook link to the Machinery's Handbook. I will probably end up ordering a hardcopy of the book. I find it easier to navigate huge reference book like this when they are a hard copy. Still haven't figured out a way to not get lost in large PDF documents that are not meant to be read sequentially from start to finish.

Wouldn't it be nice to simply open up some book to the section about whatever your current project is to find a step by step guide. But then again, where is the fun in that. LOL

Time for this girl to head to bed and rest my overworked brain cells. Good night!

Jayne

Here is a bolt together machine, to give you an idea that simple can get the job done, there v-rail system is not that great, the general build is ok for machine like this. there are 2 video's of this build, that may help you with your design.

[JayneV]

Hi Mactec, That machine is huge!!!! I only quickly scanned through the video, I'll watch it all when I get a chance.

I've been working on the gantry beam and think I have found a good compromise between ease of construction and strength and using some of the metal I already have. It's a built up box section using the Al C-section I have and 160x12 Al flat bar. It will all be bolted and glued together using a glue like Techniglue R15, which Peter mentioned in an earlier post. I found a local supplier about half an hour from home which sells this glue at a reasonable price. Any thoughts?

[peteeng]

Hi Jayne - That's looking solid. Peter

edit - if you extend the bottom channel out you can mount the drive nut directly to the gantry vs a loadpath through 2 or 3 parts... The shortest and least parts involved in loadpaths the better. Peter

[JayneV]

That's a great idea Peter, I like the simplicity. I don't have enough C-channel to do that, the length I have is just enough for the beam, but that is not a big deal, I'll just buy a length and use what I have elsewhere.

[mactec54]

Hi Jayne - That's looking solid. Peter

edit - if you extend the bottom channel out you can mount the drive nut directly to the gantry vs a loadpath through 2 or 3 parts... The shortest and least parts involved in loadpaths the better. Peter

JayneV still needs the linear Bearing mounting plate, that plate could extend out and do the same thing, it also still needs the gusset end plate

[mactec54]

Hi Mactec, That machine is huge!!!! I only quickly scanned through the video, I'll watch it all when I get a chance.

I've been working on the gantry beam and think I have found a good compromise between ease of construction and strength and using some of the metal I already have. It's a built up box section using the Al C-section I have and 160x12 Al flat bar. It will all be bolted and glued together using a glue like Techniglue R15, which Peter mentioned in an earlier post. I found a local supplier about half an hour from home which sells this glue at a reasonable price. Any thoughts?

This is just a full sheet machine, and was just a look see, at the bolting together of the main frame, yes using epoxy and a bolted joints should work well, if the surface prep is good, there is more than one brand of adhesive that will work, some being more convenient to use, here is a 3M guide for adhesives

[peteeng]

Hi Jayne - Now you have a construction philosophy figured out you may consider using a single component adhesive vs the epoxy. Loctite retaining compound 638 is similar shear strength to the R15 (and other epoxies ) and is easy to use. Plus its very good for thin film applications. The tendency with the putty is to squeeze it all out under bolt/rivet loads. The retaining compound is easy to use and easy to clean up on the edges. Surface prep same, acetone, roughen 120-240 grit acetone then apply... but follow suppliers instructions. Peter

[JayneV]

Hi Peter, Thank you for the additional info. I was also looking at 3M DP420. The 3M product has glass micro beads mixed in that claims to prevent being able to squeeze out too much glue and maintain the optimum adhesion thickness. There are so many products out there.

I've been working on the saddle and Z axis design. It's all starting to come together. Here is a progress screen shot of my progress so far. I'm modifying the previous design so still need to clean up the drawing but hopefully you can see where I am heading. The Z bearing carriages are mounted on the saddle. It's an option I am exploring to see if I can make it work.

Jayne

[peteeng]

Hi Jayne - coming along - More suggestions:
1) The Z axis plate will need to be thicker or have webs (flanges) that come out quite a bit to improve the bending stiffness of the plate. The spindle can live inside these comfortably. When its at the bottom of its travel there is a very large bending moment on it and it will flex if too thin.
2) The saddle as usual is very congested with cars and bolts. See attached sketch. If you move the top gantry rail to the top back and the drive screw to the top you can minimise the forward stack height of the Z axis (the drive nut "drives" this height). Removing the fwd spacers. Plus it gets the bearings further apart for better stability and improves access to the car bolts.
3) get a Z axis motor with a brake this will stop it falling when there is no power on, having fun yet? Peter

[peteeng]

Hi Jayne - so the next question is - Is this going to be A) a bolted or rivetted construction, no adhesive needed, the friction co-efficient of Al is very high so clamping forces are taken full advantage of and it can be dis-assembled) B) a bolted construction using adhesive as a supplement C) the structure is bonded together and the bolts are construction aids not structural? This will determine the adhesive you use to some degree...Peter

[JayneV]

Hi Peter, thank you for all the suggestions.

1) increasing the Z plate thickness and adding webs/flanges are items on my to do list. The plan is as you suggested for the spindle to sit in the space between the flanges.

2) the overlapping X and Z ballscrew assemblies have been pushing the forward stack height further out than I liked. I almost gave up trying to make this arrangement work but decided to keep going hoping the solution would present itself along the way. There is no attached sketch to your post. I'm not sure that I follow what you mean. Do you mean relocating the top X rail to the top of the gantry beam and the X ballscrew also to the top of the beam? See sketch below. That would place the X rails in two different planes. I like the idea but I am concerned that might make assembly more difficult to ensure the carriages run true and don't bind. If the rails are installed in the same plane, alignment in that plane will take care of itself. Having said that, I really like the idea and will amend the drawing to explore that option.

3) reading your YaG build thread and other build threads, a Z brake sounds essential with a heavy Z axis using a ballscrew. Are there stepper motors with inbuilt brakes, or is it an electromagnetic brake created by shorting out the motor windings when power is removed?

As a matter of fact, I am having a lot of fun.

The construction is intended to all be bolted with the exception of the built up gantry box beam. The beam will be bolted and glued. I want the individual beam components to behave as one homogenous part as much as practical. Do you think bolts alone would be sufficient? I am also considering riveting the gantry beam together instead of bolts, but that might be uneconomical by the time I purchase structural rivets and the tooling to set the rivets.

Jayne

[peteeng]

Hi Jayne - sorry about the exclusion of the sketch but seems you figured right. The concern is good but the easy path is sometimes not the best path. Being a bolted and glued assembly means things can be stacked up, checked they work as intended pulled down and rebuilt and bonded and yes it could be completely fastened just like the Eiffel Tower and Harbour Bridge. Just need bolts in the right places. You can also use a wicking adhesive to set the joints if you don't want to pull it down. Look up Loctite 290. This also means it can be pulled down in the future.

Steppers do come with built in brakes - look up stepper on line I think your pace will pick up now the design fog is clearing... Peter

[JayneV]

Hi Jayne - sorry about the exclusion of the sketch but seems you figured right. The concern is good but the easy path is sometimes not the best path. Being a bolted and glued assembly means things can be stacked up, checked they work as intended pulled down and rebuilt and bonded and yes it could be completely fastened just like the Eiffel Tower and Harbour Bridge. Just need bolts in the right places. You can also use a wicking adhesive to set the joints if you don't want to pull it down. Look up Loctite 290. This also means it can be pulled down in the future.

Steppers do come with built in brakes - look up stepper on line I think your pace will pick up now the design fog is clearing... Peter

Peter, I really like this idea for the X rails. Correct me if I'm wrong, I think this would also help distribute the beam torsional load through the fasteners more evenly?

[peteeng]

Hi Jayne - which idea? I presume the diagonally mounted rails. Its not the torsional load, its the local direct load. The rails being further apart means the spanner is bigger so the local loads are down. But it also uncouples the congestion of the saddle and the fwd stack distance. Here's a calculation showing what the break away strength of a M8 C8.8 screw is vs adhesive. There is also a more technical reason and that is the action centre of the torque is closer to the shear center of the beam of you want to look those things up... but its all up to you to move fwd on these things. Peter

[peteeng]

Hi Jayne - to discuss the fasteners. If a bolt or fastener is in a clearance hole then its the friction at the faying surfaces that transfer the shear loads. The bolt also transfers traction loads and compression loads don't need the fastener.

If the fastener is a snug fastener and not done up tightly then the load is transferred mechanically through the fastener to the hole sides. This type of connection is achieved with reamed holes and precision screws often firm fit. Sometimes its also done with dowels and bolts.

Generally in your structure you will be using clearance holes and friction type connections or adhesives. Peter

[JayneV]

Hi Peter,
Yes, I was referring to the diagonally mounted rails. I've started reworking the design to have the rails mounted this way to better visualise how it will all fit together. Only problem is I've messed something up in the project timeline and fusion keeps crashing when I make a change. I made a new copy of the model every time I made a major change to the design, and also have the automatic version history Fusion creates whenever the file is saved, so it's not a huge problem, but the crash happens when I try to delete a component that was created a little way back in the timeline, so I'd rather not have to rewind that far back unless absolutely necessary. I'll post updated CAD images after I sort out this minor hiccup.

I made the attached sketch to see if I have understood you right with your last statement about the action centre of the torque. The left sketch represents the diagonally mounted rails and the right sketch shows the front face mounted rails. The crosshairs show where the centre of torque would be in each case. This is obviously very simplified and is only considering the static weight load of the Z overhang acting on the beam with the beam fixed in space. From this sketch, it's clear that in the left case the moment arm from the centre of torque to each rail is longer, resulting in lower local forces at each rail than in the right hand case for the same torque value. And the centre of torque is also better situated in the centre of the beam in the left case. Did I understand all that correctly?

Thank you for the screw vs adhesive comparison. The calculations are all a little beyond my understanding but I will do some research and work through them to learn what all the figures mean.

Jayne

[JayneV]

Hi Jayne - to discuss the fasteners. If a bolt or fastener is in a clearance hole then its the friction at the faying surfaces that transfer the shear loads. The bolt also transfers traction loads and compression loads don't need the fastener.

If the fastener is a snug fastener and not done up tightly then the load is transferred mechanically through the fastener to the hole sides. This type of connection is achieved with reamed holes and precision screws often firm fit. Sometimes its also done with dowels and bolts.

Generally in your structure you will be using clearance holes and friction type connections or adhesives. Peter

That all makes perfect sense. Clearance holes is what I had in mind with my available tools and skills with adhesive supplement, as previously discussed, for the gantry beam subassembly.

[peteeng]

Hi Jayne - Your sketch re action centre and shear center are correct. The shear centre of a closed section is the centroid. In an open section like a channel the shear centre is not the centroid. The shear centre is the place that if you apply a load the section does not change shape or twist. I have modelled rails in many configurations and this effect is small, I think because closed gantry shapes are used. I think minimising the fwd stack distance and decluttering the saddle has more effect then the shear centre but its good to know. A good design is a summation of lots of good details.and a boiling down to as simple a design as possible. Peter


https://theconstructor.org/structura...r-center/3677/

Theory | C3.2 Shear Centre | Solid Mechanics II

My other thought for you is to not invest too much time in sorting the cabling out. I find this difficult to do in CAD. Every machine has turned out different then I expected with cable runs and placement. So think about it but it is easier to sort in the flesh once the structure exists.

Re CAD models - they do get cranky over time unless you are very disciplined with its build. Be prepared to rebuild occasionally nearly from scratch to correct issues. Build in clear sub assemblies and use midplanes for alignment vs features. Some features will change during development and this will create internal issues with the constraints. I use Alibre, haven't used fusion but I'm sure the issues are the same as models get complex...

[JayneV]

Hi Peter,

The links to the solid mechanics theory re shear centre were both very interesting. If you keep this up, I will be inspired to return to uni to finish my degree.

I think I found the problem in my model that caused fusion to keep crashing. There was an empty copy of the Z axis subassembly that I could not delete. It was empty because the assembly had no components in it, the only thing in the subassembly was XYZ coordinate reference which is somehow being referenced by another feature within the model, so trying to delete it was causing an unrecoverable error. My solution was a workaround to simply hide that empty subassembly from the timeline. I have been trying to keep the model organised by using sub assemblies and sometimes nested subassemblies. Having a good workflow in CAD is a learning curve in itself. With each iteration of the design I learn more and more about how fusion works. I’ve found it helpful when the timeline history grows too much, to disable the timeline and then tenable it. That seems to do a reset of sorts, converting many individual steps into a single base model for each component. I don’t know if that is the preferred way if doing things but it seems to work for me. Performing the “reset” also speeds up the processing time on my ageing laptop.

Getting back to the CNC, I am working on a few alternative arrangements for the X rails and ballscrew. I’ll try to model the different options today and post some screenshots.

Regarding the cabling, I’m including the cable chain in the design so that I don’t forget to allow a path for the cables. It’s not exact, more of an overview of how I think it will be positioned. In a very early version, I had completely forgotten to consider the cabling which would have then required a whole auxiliary structure just for the cable runs. The cable chain model is one I downloaded from grabcad. The model only included the individual links which I made multiple copies of to form the chain. It was kind of fun to join the links with motion constraints and then drag one end of the chain around the screen and watch the whole thing animate like a drunken snake. Lol What can I say, I was bored.