In-situ monitoring options 8

[azieg]

Hello. I was curious what most people are using for their in-situ monitoring of their AM processes. If anything. There are a couple of products on the market, and was curious of the pros and cons of each. Seems this is an area that is still being developed. Any help would be appreciated. Thanks.

 

[MacGyverS2000]

Webcams... use 'em for AM, lasering, CNC'ing, etc. I'm never far away, though... usually sitting right beside the machine, but watching a small window on the monitor is easier on the workflow than standing over the machine watching through the window for the entire run.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[JNieman]

Webcams are pretty popular.

Monitoring CNC machine codes is done with a plethora of different software options, some incorporated in the machine controller, some remotely monitoring certain sensor states or controller states. For AM, I would think the webcam is all you need, but being able to remotely monitor some conditions has benefits.

 

[arunmrao]

Webcam is what I have used and works fine for me.

"Even,if you are a minority of one, truth is the truth."

Mahatma Gandhi.

 

[azieg]

Thanks for everyone's input and apologize for my naivete in advance as I'm new to this. I suppose the webcam is monitoring the meltpool for a laser powder bed? For parts that are quite large and are safety critical, such as for aerospace, are you literally watching the monitor that the webcam is hooked up to? Seems time intensive for parts that take a while to make. Are there any tools that can automatically kill the process if certain deviations are detected?

 

[JNieman]

Parts that are more critical typically have some form of nondestructive testing to validate the quality of the part.

The answer is "yes" to all of your questions, followed with "depending on the details"

Yes, a 'shut down' can be issued if certain parameters are exceeded on any particular sensor or error code. That's pretty commonplace and standard. You can watch the monitor for FDM machines, for example, to ensure something doesn't go haywire - part shifts, nozzle blockage, or your support structure was inadequate and the part is drooping, etc. There isn't -much- you can do (within practical means) for those situations, as the machine is going to blindly keep-on-truckin' since it senses no problems.

With experience, vetted processes, and proper machine care and maintenance, you don't need to monitor things AS MUCH.

The requirements and solutions will vary based on machine type and typical part run time, I expect.

 

[itsmoked]

I recall there is is a machine that measures the part as it is being formed and will stop the moment it goes out of the spec you've set. But that is a specific 3D printer.

There is also a 3D printer controller you can use on many different machines that connects via the internet and shows you what's happening and allows you control of the printer. It's about 200 bucks IRC.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Sparweb]

We have Stratasys software that includes a "print queue" interface. You can monitor print status from a computer anywhere on your network with it. Our FDM machine does not currently have real-time feedback or a camera, as suggested above, but the print queue does allow you to monitor the progress of each print and subsequent batches, if that's all you require.

STF

 

[EdStainless]

But none of the production machines run in closed loop control.
They don't use feedback to keep welt pool, heat rate, cooling rate, scan speed, and a few other variables in a control window.
Right now it is considered 'exotic' to just watch and record where in the build there was an 'event'. Then afterward they try to figure out if it mattered.
Considering that no one in production is currently designing the microstructure there is a long ways to go.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Sparweb]

Feedback is not necessary as long as there is a record to review after the parts are produced. Component dimensions change as they cool, so I wouldn't want to try to take measurements until it had been conditioned after coming out of the printer.

"Design the microstructure"... Heck, can you imagine how anyone would inspect the microstructure?

It is issues like these that make me struggle to see how 3D printed parts could be approved for structural use in aircraft.

STF

 

[arunmrao]

"It is issues like these that make me struggle to see how 3D printed parts could be approved for structural use in aircraft."

I have been skeptical of this aspect, but since the Gurus suggest successful replacement of legacy parts by 3D printed parts, I tend to agree.

The present race is towards my 3D print looks better than yours or the least more accurate.

I have been able to accept 3D printed sand moulds and cores or wax patterns for investment casting.

But, metal printing I am still learning to come to terms. Hope, someday, some of my concerns will get answered.

"Even,if you are a minority of one, truth is the truth."

Mahatma Gandhi.

 

[azieg]

Really interesting discussion. I appreciate everyone's feedback. Especially the note about dimension metrics and cooling. With all the hype around in-process monitoring, I wonder if only a handful of variables are able to be monitored because of this. After that, I suspect CT scans for QA of metal parts are the next best thing. I'm curious for metals how different the dimensions are during the cooling process (e.g., t=0, t=30sec, t=60 sec, etc.).

 

[JNieman]

When quality requirements reach a certain point, 3d printed parts are inspected just as you would castings, to ensure expected properties are held. This might be x-ray, mag particle, or other analysis processes - I'm not as experienced with that. So long as you screen out the nonconforming product, there's no reason to exclude -any- process.

 

[EdStainless]

The way that they are using AM parts now is either they have very low loads (fuel nozzles) or they are cutting the SF from 50 to 20 and saving a lot of metal on a part that is still way over designed.
When a part is cast for forged we know from practice what the structure will be, and we deliberately modify the process (or post treatment) in order to control the structure, and achieve desired strength and fatigue properties.
Right now with AM you don't know any of those properties until you make a part and destructively test it.
This is one reason why AM wax or sand work has been so valuable, we know how to get the desired properties from that point.
There are people working on controlling heating rates, deposition, and cooling rates in order to control microstructures. This will part of the key. The other part is detecting deviations of that those areas can be tested.
Today high res CT on a metal part costs more than building it did.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube