Thread: Bespoke CNC
I manufacture aircraft propellers. The shape is machined from laminated lumps of beech by a company I sub-contract for the job. I've decided to purchase my own machine to bring this operation in-house. I've been looking around at 'off the shelf' machines but they don't seem to be so appropriate for very specific purpose and I will be buying a lot of features I won't use.
I need a very solidly built machine to minimise resonance with a working area of around 2.5 m x 250 mm and about 200 mm travel on the Z-axis.
I was tempted to design and build my own machine, the resources seem readily available and I have a mechanical/electrical background but I could easily spend all my time building and not getting on with production.
So....I'm looking for a company who can build me a machine to my specific specifications, nothing complicated about what I want, just sized and shaped for my application.
Thanks, Prop Man.
Are all the propellers the same, who currently does the cad/cam
Just learning the software can take a long time, you would need 4 axis and the software would cost around £5K +
What is the tolerance on the props, how many do you need to make.
When/if you make them yourself and you screw up YOU PAY.
I'm sure a lot of peeps on here could make this, me included but don't know a dedicated company that makes them.
Thanks for your reply Phil. It is very unusual to make 2 propellers the same. We make 2-3 a week and business is growing. I design the propellers mathematically then I draw them using Rhino. Then I pass the 3D drawing and the lump of wood to the machinists. I have looked over the shoulder of the company doing my machining and the CAM software doesn't look that complicated, the sort of thing I'm sure I can learn. At the moment I am spending about £25K a year on machining, it certainly makes commercial sense to gain such capabilities as it is a major part of our manufacturing process.
I'll be happy with a tolerance of 0.2 mm.
At the moment they are being made on a large 3-axis Thermwood, which seems to do the job well but the job is taking about 6 hours per prop including programming. I believe the machining time can be speeded up considerably with better tooling and fixtures. The loads on the cutter are quite high and the tool has to be extended from the holder by about 150 mm in some cases to gain clearance from the work. I'm supposing that a 3-axis machine will have better rigidity at the spindle, for such big heavy cuts when roughing out, on a pretty hard wood. I'm looking for a 20 mm tool holder, again to help minimise resonance.
How do you think a 4th axis will help the job?
Keen to enter discussions with anyone with track record building large-ish machines who can take on this project.....
4th axis means a rotational axis, and it would seem to be impossible to do this without a rotary ?
Heavy cutting means potential for movement/deflection of the prop
Are they doing this with one cutter setup, this is where expensive software can cut down machining time as well
Maybe post a sample file to show the size/shape of what you are after
Do you mean rotation of the work for the 4th axis? If so, the work is turned over manually. The propeller blades do suffer from resonance when the finishing cut is being performed on the second side as there is little material left. The work is held in the centre and at each end by the parent material, which is cut away once the process is finished. In addition, wooden wedges are jammed between the bed and the work, about half-way down the blade for the final cut on the second side to add rigidity and improve the finish.
I have some good ideas of how to build a much improved fixture system which allows the work to be well supported and rotated manually along it's long axis to machine the second side. Fixture building is something I will do as I have the capabilities and experience to do this.
They are roughing it with a 25 mm braised tipped square cutter and finishing with a replaceable HSS bladed cutter with a radius on the corners.
I'll try to work out how to put some pictures up. Propellers range from around 50" long to 95" long. The blank starts off as a square section and most of the material is removed during machining.
Pictures of the Thermwood cutting a propeller in the 'galery - albums' section.
It depends how much money you want to spend.
Due to the resonance/flex issues, I'd say a normal 4th axis isn't going to be much use, however a tilting table would be, as it would allow for faster maching. As instead of the existing method whereby a radius cutter works side to side the length of the prop, with a tilting table, a larger flat sided cutter can make several passes from root to top to provide a smoother profile. However, it will add alot of cost, and complexity to the machine.
I'm often surprised at the lack of companies offering bespoke machine building, but then there is the issue of who becomes liable when things don't work as intended, which means such companies would probably start charging an unaffordable premium to allow for having to redo things when they don't quite work as initially planned.
Haven't checked out the pics yet but could you not still manually turn the piece for each side but add tabs on the blade edges? it would stay rigid in the block then. These have to be cleaned up anyway to balance the prop. I would go with a KISS approach my self?If the nagging gets really bad......Get a bigger shed:naughty:
Thing with bespoke machinery is that A) frequently the cost would be prohibitive, and B) if you go to sell it in a few years to upgrade, it may not easily sell (unless you sell it to a company doing the same kind of work).
You sending orders direct on Rhino / CAD / CAM?
With one-off jobs, 40% of your time can be on set ups (clamping / clocking / changing cutters etc). Bear in mind with wood you can't use coolants, so feeds and speeds can't be ramped up quite so high as on other jobs, and along with that, there is the issue of tool wear as well, and that'll be high.
How are they programming the machine? Directly from CADCAM, or G-code from the control panel?
Props certainly don't strike as an easy job, so 6 hours per job sounds like a pretty good time from here, but I haven't seen the job being done.