Nottingham engineers are pioneering a waterjet milling technique that can cut at precisely controlled depths and self-correct to produce complex geometries.
The subtractive manufacturing technique is particularly useful for very hard materials such as ceramics and is being used for aerospace applications and medical prosthetics.
While waterjets have been used for decades to slice clean through sheet materials, the ability to mill shapes with jets is relatively new.
The key to enabling this is the ability to control the depth of penetration of the jet into the part being milled.
‘It depends how much time we’re exposing the part to the jet — so we’re moving with different speeds,’ project lead Prof Dragos Axinte of Nottingham University explained.
‘We developed some models and we know how the part will erode at particular feed speeds; that’s the key element behind the work we are doing.’
The team uses water laced with abrasive garnet particles, which is forced out under pressure as a jet 1mm in diameter travelling at around two to three times the speed of sound.
Another key element to the project is the ability of the jet to self-correct in real time as it is moving along — although this presented a particular challenge for the team.
‘If you imagine a jet with a high velocity going into a surface, the reflection is so bad you cannot see anything. It’s wet, misty and full of grit that’s bounced back. You cannot use a vision sensor or put a force sensor in, so we developed a monitoring system based on acoustic emission sensors,’ Axinte said.
The sensors are calibrated beforehand so they can match an acoustic signature to the status of the surface. The self-adaptive module then makes the necessary adjustments.
‘It is a niche technology; it can be economically viable only when you cannot cut with other processes — so with very hard materials such as ceramics, or materials that you do not want to thermally affect such as nickel and titanium alloys, also parts you don’t want to deflect because actually water machining results in very low cutting forces, less than 5N,’ Axinte said.
The €3.8m (£3.17m) EU FP7 ConforM-Jet project is led by Nottingham University in collaboration with researchers in institutions across Europe, including Zeeko (UK), BAE Systems (UK), Finecut (Sweden) and the Royal Institute of Technology (KTH; Sweden).
This is a very interesting machining technology. With respect to the self-adjusting jet contact-time with the object I would conclude that the abrasion rate is a complex function of the shape of the surface that is to be machined and not just a factor of exposure time of the jet at a particular spot. For example if material is to be removed from a rather deep narrow cavity the reaction force of the jet is much larger than if the material is to be removed from a flat surface, or from a surface that is at an angle to the jet. This is a simple function of the momentum change of the fluid and the entrained grit. It would appear interesting to relate material removal rate with the orthogonal reaction force to the jet and it angle with respect to the surface to be machined. . .this might be a more efficient feedback parameter for the “cutting” than simple jet-time exposure on a specific location.
Vortex Engineering
Nieuwegein
Interesting article – although the idea that you can ‘mill’ with a waterjet is not at all new. We can already do it in 2D – its just very expensive and therefore not widely used.
However the idea of acoustic driven self adjusting jet contact time is pioneering and if workable would move the process forward. (There may be potential to utilise this technology to monitopr all waterjet for jet quality) The next step to take it to 5 axis would be truly progressive. However the key is the quote about it only being applicable to a niche market due to the economics ie its still going to be very expensive and therefore a last resort?
Just to mention this is done on a 5-axis machine; the control strategy for generating freeforms can also employ tilted jets.
The technology can be considered “niche” mainly because it can be successful where the others (conventional machining) might not be economical. For example, we generated pockets in ceramic composites that might be otherwise difficult to machine/grind.
Just to clarify:
– the material removal model predicts the jet penetration for a particular set of abrasive parameters (e.g. pump pressure, abrasive grit flow) and feed speed, tilt and direction of the jet
– the acoustic emission mounted under the sample in the fixture detects the level of material removal: if this is constant (as predicted by the model) no action is taken; if there are deviations from the programmed jet penetration then the signal will allow to adjust the jet feed speed.
Note we tried several of other sensors but the harshness of the work environment does not allow this.
In my remark on April 5th I suggested that the “force feedback” method is a simple matter of the “momentum change equations”. . .this was not intended to suggest that we think that the solution for accurate feed-forward material abrasion rate function would be simple :-). It means about the same as to say “fluid dynamics is a simple matter of applying the Navier Stokes Equations, which are notoriously difficult to solve.
It appears that already serious effort with complex methods has been applies to predict the abrasion rate. Congratulations!