Printable View
Hi Mactec - Looks like its the LR5201's they are $24AUD local $19 from Vxb so no gain there and the idler bearings local are $40 each and narrower then the 5201's. So its 5201's and a bolt. I can get the 5201 for $16AUD each on ebay free freight to my door... Onward cheers Peter S
Hi peteeng,
I am catching up with the Maximus. Nice read! (not finished yet)
I was wondering, did you consider using a belt reductions rather than gear boxes ? What would be the main tradeoff's ?
Cheers,
Hi Planck - it costs more to make a bespoke gearbox. The gearbox from the motor people costs $58AUD. I doubt I can make a 5:1 belt reduction for that amount, certainly not a 10:1 as the pulley size is huge or have to do a double reduction then its certainly too many $$$. Remember I'm making this to resell so every $ spent has to make sense.
Update - I've modelled the cam follower bearing on the return pulley and found that the bolt head is buried so I can't get at it to adjust it. So looks like its two deep groove bearings with a spacer. I may look at moving the belt up a bit so I can get a key under it but that's a fiddle.
More update- I've decide to stay with the cam follower and use a spacer. It's always a fiddle getting two bearings and a spacer to play well together. Did anyone get on their school reports "does not play with others well? mine was, can do better" Using the single bearing means this issue is removed. cheers
It's also very difficult to make a single belt reduction with a ratio higher than 3:1. Due to the difference in pulley sizes, you don't get much engagement on the small pulley.
Hi All - Fitted the spacer and washers to one of the return pulleys today, looks good. Now to detail the tensioner and that part of the puzzle is done. Then onto the cross gantry mounts and most of the small items are off the list. Still thinking about the 32mm belt... Peter
Hi All - Rebuilt the cross drive anchors and got word from the bender that the 45deg bend is fine. Now looking at the cross drive return end, I see I can't use the tensioner from the long axis so will have to think about that. The list of little things is nearly complete so can start Round 3.
The bender sent back an image from their simulation software so I attached it here. This checks that the bend happens and nothing fowls or slips during the bend.
Started thinking about bump stops but one side doesn't match the other :) Will come out in the wash...Peter
Hi All- Here's Max's first drawing!! It's like putting your kids drawing on the fridge!! Well this means the drawing stack will start rising... closer to quotes.. Peter
Hi All and Sundry - Been sorting the bench tensioner and return pulley mount. Have used a flange as a register so the mount is the correct distance from the rail. Will do something similar at motor end. The tensioner at the cross gantry belt run is tricky, no brilliant thoughts there yet. I think I'm going with the 25mm belt. I don't like the 10mm shaft overhang the 32mm would have. But I can make the mount with enough height to use the 32mm if I want in future.... not at that cross road yet. Peter
Hi Peter,
Not worth fitting an outboard bearing on the pulley to take some of the axial load off the stepper?
Hi Sterob - Currently the way the motors are mounted there is no room to create a support shaft. I'm happy with the numbers nothing will fail. I've resolved all the tensioners. Need to fix the motor mounts and this part of the puzzle is done.Then I can work through the parts and add a bit of style and Actio (as Planck would add) , then its a total rebuild into subassemblies and nuts and bolts. Peter
By the way there is no axial load only a shaft transverse moment on the gearbox. The stepper itself is well insulated from these loads via the gearbox.
Ahhh...yes...I for got about the gearbox......lolQuote:
Originally Posted by peteeng
Hello All & Sundry - This morning I tidied the motor mounts and it seems all the parts are there functionally. It's time to retire the Round 2 or P2 model. There are 357 constraints in it and most have fallen over!! That's the way these development models go, its not worth patching and nursing R2 any longer. So now all the parts and their relationships exist I can start building R3 in a clean and organised manner suitable for manufacturing documentation. This also means doing a drawing for each part and assembly. The workload only increases in this part of the project and this is a small machine. The drawing stack can be a seriously demoralising slog in a big project. It's worth having a party once the stack is full. In the current P2 model there are 54 individual parts but the part number is up to 249-140. This means I have designed 140 parts and used 54. About 40% retention rate not bad for this sort of project. So a quick review to check I have not missed something then it's time to start making drawings. The project now will be driven from the drawings. The models are related to the drawings and the drawings hold all the manufacturing information. It's great to have solid models of things, but tolerances, specifications, materials, relationships and processes are all captured via the drawing. Once the drawing is established you find the model via the drawing to ensure you use the right one!! Then we get into change management practices with issue or revision systems. Every company I have worked for, these have been different and complicated. I keep it very simple. That's another huge topic... Regards Peter
Edit - Waterfalls - I have been working on an architectural aspect to add to the design. I like art-deco particularly "streamlined" if anyone's into this sort of thing. Streamlined buildings had a feature called a "waterfall'. It's a column structure on the facade. I recently was in Napier NZ and they have many art-deco Buildings. So I have added some curves to the parts. What do the viewers think? I suppose they could also be called buttresses. I could make them into flying buttresses...
nice.............:)
Hi All & Sundry - Being an engineer I suppose they are buttresses. Tidied up a few rads and made some curves consistent. Peter
Hi All & Sundry's - It's time to review Round 2. Maybe even comment on the whole lot. I started on the 14-3-19 and it's now 06-06-19 So its 3 months from the napkin sketch to a functional design. The original concept was to use a corrugated plate for the Z axis and this has dissolved or morphed into something else. The foundation bubble of the design was to have a 250mm Z travel and be able to cut aluminium. The issue at the time was how stiff does this have to be? I didn't know but the forum answered that with the attached thesis and some other info. So now we had a static machine stiffness target of 10N/um to chase. This is the stiffness of a good small mill. I also designed this from the Z axis out. I and others usually design from the bench to the tool. Going from the tool outward meant that I didn't know the size of anything really until I got to the end. But the reasoning was that in this method each part will be its correct size for its stiffness vs the other way in which compromises are always made along the way to squeeze things into the ever shrinking design space.
The other foundation idea is to minimise the tool overhang or the forward stack height of the tool from the gantry. This was part of the reason for the corrugated tool plate. The tool overhang is quite large on many machines and any cantilever is a vibration and deflection issue. This resulted in the side drive being developed.
Sorry can't attach the thesis so search for Principles of rapid machine design - Eberhard Bamberg
The rough footprint of the machine is 3m x 1.8m and will take a full imperial sheet of 8x4ft with a 100mm margin around its edge. I checked the shopbot 8x4 machine and its footprint is 3mx1.975m so I'm not so worried now, thought it was going to be very wide. Their motors stick out of the sides . AvidCNC 8x4 footprint is 2.9m x1.85m so close to Maximus. I'll recheck the width...
The bench is to be built by the Maker in metal or timber. The design is intended as a kit machine. An economy less stiff machine is envisaged once the heavy is resolved using the ame design parts but thinner. R1 & R2 where very much about form finding of the parts and resolving the topology of the machine. The parts need to be manufacturable and scalable particularly by using laser cut parts. They do show some heritage of their predecessors, Scoot A, Scoot B and Brevis.
The machine is developed using an FEA approach called BVA(boundary value analysis). This is via a program called Simsolid. It can analyse very large assemblies down to the nuts and bolts level quite quickly compared to conventional FEA. The CAD side is handled by Alibre a parametric solid modeller for mechanical design. And the brain power is supplied by my morning muesli and coffee. Right now we are at the drawing and manufacturing detail level Round 3. Then a final round of structural simulation to tune up the design to determine its real stiffness and onto what is officially called "polishing". A polished design is always best... So we have travelled from the fuzzy front end to the middle of the project.
There has been some twists and turns and some elements of the design are unexpected, eg the tubular tool plate, but that's the fun of pursuing this sort of project. Thanks for watching and participating. Peter
Edit - Future thoughts
1) The gantry cross drive would be good to be a ballscrew. Perhaps the long axis too.
2) Will pursue finding the ATL belt as this is the best belt for the job
3) Haven't found a suitable part for 3D printing or composite fabrication. Will look at this again in R3. The obvious thing is to make a carbon fibre gantry but this is a big part at this point in the project
4) Need to look at a metal bench, a basic design would be good
5) I'm thinking about putting the belts on the inside of the column to gain a little more cross travel, will look at the numbers to see if this is worthwhile
6) I'd like to connect the column brackets together somehow but no brain waves on that yet
The intent is to make a kit that never needs to be structurally upgraded and its stiffness will never be questioned within in its class. The drives and other things can be tinkered with and the DNA of the structure will carry this through. Onward to Round3...
Hi Peteeng,
Gosh, if ever you come to Europe you should definitively visit Prague in the Cezch republic (maybe you did already). One of the lesser known birth place of Art nouveau and Art deco, and some exceptional craftwork and architecture. There is also a nice technical museum that will surely summon your mechanical engineer self. They had quite a nice industry back in the days. From what I know a key characteristic of art deco design is in mixing old skills and new techniques. Yours buttress are looking spot on.Quote:
Originally Posted by peteeng
About the hole project, I really love your conception process. Kind of an applied retrograd analysis one could say. Unfortunately I can't comment on the technical aspects, you certainly know your business better than I. All I can add is that as an end user I would strongly consider the machine accommodation (I hope this makes sense) , such as the ease of exhaust system integration or compactness of the hardware material. Another important factor would be readiness of maintenance.
I am looking forward for the R3 report!
Cheers,
Hi All & Sundry - A bit of project trivia. This morning I have loaded the 2019 Alibre Beta. It has improved video sharpness which makes looking at the work much easier and pleasant. It also has an illustration setting which I quite like for documentation and will use it on the webpage I am building at present for Tetra. cncrouterkits.com.au
Hi Planck, thanks for the encouragement. Although not discussed dust control, maintenance and various operational stuff is in my head and slowly its being integrated into the design. Maybe in the polishing stage these will be more strongly represented. Being a kit its up to the Maker to deal with many of these aspects. But I shall be putting in extra holes and flanges in spots to accommodate that sort of thing. A few extra holes don't cost much and its surprising how handy they are. Also being a regular user and abuser of these machines gives me a perspective for this sort of thing.
I was in Paris a few years ago and stayed in a wonderful hotel with lots of brass curvy bits. The artisan work was exquisite as is the subway.
Your "hole" project threw me for a bit then I realised you meant "whole" :) My wife is an ex english as a second language teacher and we laugh about that sort of stuff often. :) Peter
English is hard on me .... And good to know that you are thinking about adding some wholes here and there.
Hi All - I have found an australian source for the ATL belts. But they are pricey. The AT10- 25mm costs about $35/m the 32mm about $46. The ATL10-25mm is $48 and the 32mm is nearly $80/m. Seems the ATL10-25mm is the sweet spot. The ATL is 1.57x stiffer then the AT. Info is good.
The AT10-32 is 32/25= 1.28x stiffer then the AT10-25mm
The ATL10-25 is 1.57x stiffer so the AT10-32 and the ATL10-25mm are about the same price yet the ATL10-25 is a bit stiffer.
Plus consider the ATL is profile and length compensated for higher preload....Peter
Hi All+Sundry - I have started setting up the drawing system. The bolts and nuts are always an issue. If I put them into the subassemblies I have to manually compile them for orders which is a pain. If I put them into the top assembly they don't appear when you look at the subs :) So I think I have to go the usual way and put them in the subs. I'll try an "atomic BOM" at the end to see if this works. An atomic BOM is a BOM that includes everything in one list, then I can extract the nuts &bolts as one list. This means changing some things in the subs so this can work. Hopefully with this change it can be changed back with no consequences!! Haven't released an atomic BOM before so will be interesting...
things will be boring for a while as I do dwgs and organise the model... detail takes time. I'll try to think of something to entertain you all with. Peter
I published a video of Maximus moving in CAD. Spot the intentional error
https://www.facebook.com/watch/?v=2393246720931215