Sheet metal forming – a new method

[Ches Green UKParticipant @chesgreenuk]

ROBOFORMING: The Future of Metalworking? – https://www.youtube.com/watch?v=dCXu8Ju_fdY&t=1451s

I saw this first mentioned on the Practical Machinist website. It’s basically two robotic fingers pressing the sheet, one each side, and tracing out the profile whilst applying pressure. It takes maybe 90 min to make a shape but is quicker than making dies for pressing out shapes, up to a reasonable batch size.

The workforce seems smart and enthusiastic.

The R&D bill must be high ; I haven’t yet checked if the company makes it’s money in other areas or whether this is a start-up company with this product only.

Possibly a tad expensive for the home workshop at the moment, but maybe a mini version will become available in due time.

Ches

[Ady1Participant @ady1]

Been around for a while now

The big weakness is it’s pretty slow, the big strength is it’s great for a few bespoke jobs

Handy for one-offs of kitchen tops at the luxury end of the market etc

Musk etc use giant stamping machines which crank out a part in seconds

Saw an interesting CNC sheet metal puncher a while back which basically only uses a few square punches and some round punches to machine gun any size and type of rectangles ovals etc into the sheet metal, completely eliminating the need for bespoke dies unless you need a huge production run

[Ches Green UKParticipant @chesgreenuk]

Ady1,

I got the impression from the way they were enthusing about it, it was quite a step forward. The video did show previous Patents dating back many years but it seemed no one had yet got it to work properly.

I thought St/St kitchen sinks were deep drawn…but I suppose some customers may require something a bit more exclusive 😉

Elon is now apparently using ‘air bending’ to crease the 1.8mm thick St/St panels on the Cybertruck ie the panels never bed down in to the bottom half of the die.

The square/round punch method sounds interesting. I’ve been away from the coal face of actually working for quite a while now so only keep up with the latest technologies via the internet and friends still working in industry.

Ches

[noel shelleyParticipant @noelshelley55608]

A very interesting idea even if it uses vast amounts of computing power and a great deal of experiment to arrive at the parameters for any given thickness, position on the sheet or material. Reminds me of hydro forming ! Noel.

[Ches Green UKParticipant @chesgreenuk]

Noel,

Yes, I also thought it was an interesting way of forming sheet metal, but why go to all that bother and expense.

The company doing it (Machina Labs – https://machinalabs.ai/capabilities) doesn’t give too much away on their website as to why they have made, what seems to me, such a significant investment.

However this article does give a bit more background to the story … https://3dprint.com/303301/us-air-force-buys-machina-labs-deployable-roboforming-system/    I think it says Roboforming allows fast prototyping and small batch production. It may also offer an alternative to “…directed energy deposition (DED) and wire arc additive manufacturing (WAAM) applications.”

But I get the feeling we are not being given the full story.

Ches

[noel shelleyParticipant @noelshelley55608]

My thoughts Ches, WHY would you invest a very considerable amount in this way, The number of robot arms etc. in the workshop and mention of Inconel and titanium. I smell a rat ! Wire Arc Additive Manufacturing ( WAAM ), is that a way to blind us with science ?  Why not call it glorified MIG welding ?

At least with Hydro forming we can do it, all you need is a good boiler test pump !  Noel.

[noel shelley]

On noel shelley Said:

My thoughts Ches, WHY would you invest a very considerable amount in this way, The number of robot arms etc. in the workshop and mention of Inconel and titanium. I smell a rat ! Wire Arc Additive Manufacturing ( WAAM ), is that a way to blind us with science ?  …

The answer is these are CNC manufacturing techniques that in certain cases are cheaper than manual and semi-automatic methods.

Rather than employ an expensive skilled team of panel beaters who take forever to create a special shape, the company buys a pair of robotic arms, some hammer tooling, and a computer.  Total cost perhaps £250,000 with a 7 year life.   This compares well with a couple of employees costing pay, tax, National Insurance, pensions, training, holidays and sickness, plus admin overheads etc.  The capitation rate of an employee, i.e. his total cost to the employer, is often about double his take home pay.  Average pay of a panel beater is about £12 per hour, so two of them on the books could cost the firm about £100k per year.    If the robots do the same job as the men, the robots save £400k over 7 years.  More if the robots work round the clock all day every day.

The savings are only achieved when a smallish number of moderately complex shapes are produced from a CAD design.  Not so useful in a garage fixing damaged car bodies, nor competitive with a press set up to knock out many thousands of panels from dies.   But excellent for limited run sheet metal artifacts like custom sinks, vehicle body shapes, fairings, and ducting etc.   The advantage is no skilled hammer wielding metal-workers are needed.  The designer defines the shape with CAD and then the robots make it.

Same with WAAM, were the emphasis is on ‘Additive Manufacturing’ rather than ‘Wire Arc’.  The Wire Arc used may be MIG-like, but WAAM is an automatic process producing arbitrary shapes by building up successive layers of accurately placed blobs.  3D printing with steel rather than plastic.  Again, the advantage is WAAM doesn’t need expensive skilled metal-workers because the CAD design is sent direct to the machine.   Like robotic panel beating, also best for moderately complex one-offs and short production runs rather than repair work and mass-production.

If you want to know why firms do what they do, ask their accountant!

Dave

 

[Ches Green UKParticipant @chesgreenuk]

Dave,

Good stuff. Yes, employee costs to a company is usually 2x salary, and robots don’t go sick, get injured  etc.

I think I read the aerospace industry (military?) may have an interest in such methods, and they pay top dollar.

Also, Machina bought out the original company who came up with the process, so Machina might be taking the process to the next level and then looking themselves to sell it off at a premium.

It will be interesting to see if it gets traction within industry. I guess it is ‘watch this space’ for now.

Ches

[Author]

Posts