Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Metal Matters LMWD Printer
As per the last post in the SLM thread I discussed the Laser Metal Wire Deposition (LMWD) system I have been working on. I've since moved the laser head over to a new platform (actually the old i3 frame from some of my earlier videos) which will serve as the basis for this process. I'm hoping to keep this as an i3 based platform as it very closely resembles FDM, lowering the barriers of cost and complexity.

[Image: 197f42f0e8b238058246ea4d172a6e1e.png]

One of the major issues that still isn't entirely resolved is the wire feeding mechanism. I ended up redesigning the wire feeder to use two roller bearings centered around a single drive wheel, now lined with PE foam to better accommodate 0.1mm wire and permit slip if any obstructions should occur. Despite doubling the friction this only produced more issues as the wire no longer followed a straight path. Instead, if the wire were to jam downstream it would quickly leave the center line of the rollers and tangle itself instead of slipping against the foam.

[Image: 673d997a8db6cfe89557ebd83ebadf68.jpeg]

I had to revert to using the original design but this time with a couple tweaks. I kept the PE foam on the drive wheel as this allowed the 0.1mm wire to be used and provided slip when necessary with the larger 0.25mm wire. The wire guides / PTFE mount are now part of the tensioner arm as this provides a straight path for the wire to follow, allowing the drive wheel to slip without disrupting the wire, making jams far less frequent. For now the performance is adequate.

I still need to think of a way of reducing the wire curl as it comes off the spool. The syringe I'm now using is 23G (vs 21G previously) and provides a tighter clearance but doesn't overcome this issue. Unfortunately there needs to be enough clearance so that the wire can move freely and this permits some wire curling and therefore some inconsistency in the delivery of wire. This is specifically the case when using 0.25mm wire (23G syringe ~ ID 0.3mm), 0.1mm wire doesn't suffer this issue as it is soft but requires an even finer syringe which I do not currently have on hand. The 0.25mm wire is classified as "hard" so I may try using a soft variant with the hope that it is less prone to curling. I suspect that this will be the case but will pose other problems e.g. collapsing at the inlet of the PTFE tube.
[Image: 7c0c8c8be219076b2ca019cfee08b94b.png]

What's illustrated above is the concept for how I intend to split the beams so that wire may be fed central to the weld pool. Adjusting the prism's height allows the breadth of the two beams to be set. Admittedly, it is more efficient to knife edge the lasers and not rely on a prism but due to the lens being relatively small, and the size of each laser mount, this isn't possible at the moment. I have ordered some larger lenses and will pursue knife edging once they arrive.

I have also been working on the head in regards to wire delivery. I had originally hoped to drill a hole through the lens to help ensure the radius of curvature of the wire wasn't too tight. However, drilling through K9 glass has proven to be a challenge so I've had to opt for something a bit more pragmatic. The syringe which delivers wire and supporting copper tubing now exists entirely beneath the lens which doesn't provide much space to work within. To try and mitigate that, the PTFE tubing which carries the wire to the syringe passes through the lens mount, helping reduce the radius of curvature and in turn lower the load on the wire feeder. I've also had to use smaller copper tubing to allow the syringe to be better supported and rely on crushing the syringe in place rather than using a modified hotend nozzle as I had done previously. Although the beams are split the space between them is quite narrow.

[Image: 949d0af3146e5faa1b63d9305b7c125a.jpg]

Despite these obstacles it seems that things are coming along quite well. The characteristics of the laser spot are what I had hoped for. As the incoming beams are aligned toward the outer edges of the lens they produce a concentric profile which is concentrated around the edge, roughly mimicking the characteristics of a professional system without the need for an elaborate lens array. Utilizing the outer edges of the lens will result in lower power density due to aberration but it's important that wire does not intersect with the laser before hitting the substrate otherwise the wire will melt and ball due to surface tension. This was one of the issues with the side entry approach as since we are working with relatively low power, wire which obstructs the laser can effectively draw away enough energy to prevent an instantaneous weld from taking place, instead forming an accumulating ball of SS316 until it is large enough to transfer enough heat to form a weld. This made the previous approach very sensitive to the angle and overlap of the wire feed. To give some context, a lot of professional machines run 1Kw+ lasers but generally cater for typical welding wire sizes (0.6mm - 1.0mm).

[Image: 9cd6254b2475ac94e732f1aee15f5ffc.jpg]

At the moment I am amending the laser mount design. The most recent laser head base plate is made from PLA as I am running low on PETG/ASA, which has revealed that the lasers will canter inwards when the laser mounts are tensioned in place. This is due to the slotted guides on the base plate only being around 3mm thick, the laser mounts having their mounting screws biased towards one side, and PLA's propensity for deformation when exposed to long term strain. I am printing some new mounts now with centered mounting screws and have added some capacity for fine adjustment for the sake of beam alignment.

[Image: 878008f6f699ad07fab476de3e4fe41a.jpg]

That's all for now.

Forum Jump:

Users browsing this thread: 1 Guest(s)