UCARE Summer Symposium Presentation

Preston S. Noll Author

08/04/2020 Added

21 Plays

Description

Video presentation of UCARE research on a study of energy consumed during milling of interlayer-peened additively manufactured 316 stainless steel.

Searchable Transcript

Search:

[00:00:00.634]Hi my name is Preston Noll
[00:00:03.540]I'm a senior in mechanical engineering and I'm here to talk to you today about our UCARE research project,
[00:00:09.260]Which was the study of energy consumption during milling of interlayer ultrasonic peened stainless steel.
[00:00:16.361]We're going to start with a little bit of the motivation for this project
[00:00:19.262]then breeze through the literature review
[00:00:21.412]and talk about our objectives and plans for this experiment
[00:00:24.996]and then some results which will then be summarized.
[00:00:28.730]The overall goal of additive manufacturing and hybrid additive manufacturing
[00:00:34.915]is to print the mechanical properties desired into the parts that we are creating.
[00:00:39.348]So we combine printing with a secondary processing method
[00:00:42.382]whether it be milling or peening, which are addressed in this experiment
[00:00:45.983]to determine what properties the material that we construct will have
[00:00:50.634]by first printing it and then introducing a surface treatment
[00:00:56.902]to control the properties as we go along.
[00:00:59.251]The problem with this is we need a non-destructive method of measuring these changes
[00:01:03.052]as what we have now is milling through the part
[00:01:06.872]to understand the surface energy of each layer of the part,
[00:01:11.570]which completely destroys what we are working with
[00:01:15.103]and renders that part invalid for further use.
[00:01:18.186]In this project we are using an Optomec LENS 500 hybrid machine tool
[00:01:26.006]which combines milling and printing processes
[00:01:29.090]in order to conduct hybrid additive manufacturing
[00:01:32.389]all within one system
[00:01:34.141]and measuring the power used by the milling head
[00:01:38.792]while milling through a part.
[00:01:42.675]Through further analysis done by Dr. Sealy,
[00:01:46.908]the primary investigator of this lab,
[00:01:51.276]it has been determined that the process level spindle power
[00:01:54.544]as seen in inset (c)
[00:01:57.390]is a good indicator of the mechanical properties of the part
[00:02:01.825]because the energy required to mill through the part
[00:02:06.509]can be separated, using a power analyzer,
[00:02:09.258]from the overall power required to operate
[00:02:11.738]the Optomec machinery.
[00:02:13.574]We plan to use the curve seen here as a process signature
[00:02:17.853]in order to determine the energy of the part
[00:02:20.554]that we're milling through and to measure
[00:02:22.704]the impacts of the secondary process that we are using.
[00:02:27.105]This is further breakdown into that can be determined
[00:02:29.756]but as you can see here
[00:02:31.022]the net specific cutting energy is the important parameter
[00:02:35.043]that we use to measure the material properties
[00:02:40.002]of the section of the part that we mill through
[00:02:42.432]and hope to view changes in the part as
[00:02:46.449]we get closer to and further from a hybrid layer.
[00:02:49.218]Our hypothesis for this experiment was that
[00:02:53.318]there is variation in the energy consumed
[00:02:55.185]by the milling head during milling of
[00:02:58.802]interlayer peened samples as compared to
[00:03:00.751]wrought control samples and as-printed
[00:03:02.869]control samples because of the secondary process conducted,
[00:03:06.970]which was a peening or compression conducted
[00:03:11.054]on the part at certain intervals.
[00:03:13.234]We also tested the hypothesis that the energy
[00:03:16.127]consumed milling through any type of part will
[00:03:18.445]vary as a result of its location within the
[00:03:20.930]part that we're working with.
[00:03:22.571]As you can see here, not all of our samples
[00:03:25.802]were prepared using the Optomec system,
[00:03:27.970]we did have to fall back onto the
[00:03:30.104]Lumex AVANCE-25 powder bed fusion printers
[00:03:32.888]as there was some overlap with the scheduling
[00:03:35.254]in the lab, but these parts were additively manufactured
[00:03:38.189]out of 316 stainless steel using the print
[00:03:40.739]parameters seen here, and what's most important in the hybrid samples
[00:03:45.527]is that every 10 powder bed fusion layers, the print
[00:03:49.558]process was stopped, and the peening process
[00:03:54.583]as you can see on the bottom of the screen
[00:03:56.276]was conducted on the surface of each part
[00:03:58.444]and this was conducted to compress the surface
[00:04:00.761]and hopefully to give us distinct material properties
[00:04:05.461]that we can witness when milling.
[00:04:09.328]Again, this further describes the process of measuring
[00:04:13.912]the energy required to mill through these parts
[00:04:16.763]As you can see these are our milling process parameters
[00:04:19.747]and some parameters associated with the
[00:04:22.597]power analyzer used for data collection.
[00:04:24.648]There was some adjustment of our experimental plan
[00:04:29.031]throughout. As you can see in the top left,
[00:04:33.599]there was significant amount of tool wear occurring
[00:04:35.916]which we concluded was because we only
[00:04:38.601]milled half of the part from top to bottom,
[00:04:42.167]so as we moved towards the bottom and deeper into the part,
[00:04:47.267]significant tool wear occurred as the tool insert rubbed against
[00:04:52.169]the remaining area of the build, as you can see on the right side.
[00:04:57.252]To correct this, we adopted the so-called stairstep method,
[00:05:00.587]where we eliminated one of our milling passes
[00:05:02.589]as we moved down through the part
[00:05:04.404]in order to minimize the amount of time spent
[00:05:07.422]rubbing the vertical surface of the cutting insert
[00:05:10.206]against the remaining material.
[00:05:13.105]Again, this describes what you can see in the milling chamber.
[00:05:17.256]This is one of our finished workpieces on the left,
[00:05:21.239]and as you can see we moved from left to right
[00:05:24.876]of the workpiece and from top to bottom
[00:05:27.558]as we milled through it,
[00:05:29.541]and every milling pass removed approximately
[00:05:33.608]ten layers of print, which comes out to half a millimeter
[00:05:39.043]or 500 microns of material.
[00:05:41.777]So these are our results, we will move from
[00:05:44.995]tool wear to microhardness testing to finally
[00:05:47.827]the maps of energy consumption during
[00:05:50.462]these processes.
[00:05:52.095]As you can see here,
[00:05:55.195]as we moved through an as-printed sample we measured
[00:05:58.679]tool wear at layers 5, 10, 15, and 19, which was the end of the milling process.
[00:06:03.295]Layer 5, 10, and 15 were minimal;
[00:06:06.580]you can begin to see some edge wear that
[00:06:09.664]may be significant at layer 19,
[00:06:11.916]but these tool inserts remained fairly pristine
[00:06:16.465]as opposed to some other examples.
[00:06:19.849]This shows the end-of-process tool wear
[00:06:22.951]witnessed at the end of all of our milling tests
[00:06:27.084]and as you can see the hybrid displayed the most wear
[00:06:30.218]but even then it was an insignificant amount
[00:06:33.586]and can be eliminated from further consideration
[00:06:39.819]as a parameter for increasing energy consumed during milling.
[00:06:43.819]Here we can see the microhardness as we move
[00:06:48.054]down from the surface of the part during milling
[00:06:51.022]We measured every 500 microns or 1 milled layer
[00:06:55.695]as far as the equipment allowed.
[00:07:00.475]The conclusions as you can see here
[00:07:04.220]are that the hybrid sample is clearly harder than
[00:07:07.899]the as-printed sample, which makes sense given that
[00:07:10.934]we compressed the sample at regular intervals.
[00:07:13.117]But, the as-printed sample also oscillates more than
[00:07:16.102]the hybrid sample, which is inherent to
[00:07:19.850]additive manufacturing, that there is some variation
[00:07:22.916]in mechanical properties of the part,
[00:07:25.002]which also validates our desire to conduct this research
[00:07:28.670]as there is evidence that the internal microstructure
[00:07:32.620]of additively manufactured parts is not super consistent.
[00:07:37.120]And here is our party piece of the presentation.
[00:07:41.003]This is our net specific cutting energy map
[00:07:43.655]of a wrought sample,
[00:07:44.938]and as you can see it increases from left to right
[00:07:47.340]which is the course of the milling path across the screen,
[00:07:50.306]and from top to bottom of the sample.
[00:07:53.540]We can see that, if you can imagine this is
[00:07:58.207]the left half of the sample,
[00:08:00.158]there will be a normal-distribution type curve in
[00:08:04.337]energy consumed that peaks towards
[00:08:06.752]the middle of the sample and there
[00:08:08.903]is a general increase from top to bottom
[00:08:10.987]in the energy consumed per layer.
[00:08:13.002]This is the same map, but of an as-printed sample,
[00:08:17.488]which caused some interesting conflicts in what
[00:08:22.688]we were observing. There is a consistent chunk of layers
[00:08:28.107]until around halfway through the part
[00:08:31.140]or about 5 millimeters down from the surface
[00:08:33.523]when suddenly we had a substantial spike in energy consumed per layer.
[00:08:37.473]It does still observe the trend of increase from left to right
[00:08:42.924]which we hypothesized would continue to decrease
[00:08:46.524]as it moved from the middle toward the other edge
[00:08:49.726]of the part.
[00:08:51.726]And then again we have the same map
[00:08:54.110]but of a hybrid sample.
[00:08:56.994]This does confirm our conclusions from the wrought sample,
[00:09:01.328]which was comforting after the irregular data of the
[00:09:04.696]as-printed sample, and again you can see the trend
[00:09:07.612]from left to right and from top to bottom of the sample
[00:09:10.646]that the energy increases towards the central bottom of the part.
[00:09:14.681]Here is that data compared side by side and
[00:09:18.940]normalized for color; you can see that the hybrid and wrought
[00:09:24.419]samples displayed that trend while the as-printed
[00:09:27.667]had a top to bottom increase in energy consumed.
[00:09:31.454]But, all of them showed a general trend to increase
[00:09:38.405]in net specific cutting energy per layer
[00:09:41.889]as the distance from the surface increased,
[00:09:44.104]which was an encouraging conclusion for us.
[00:09:46.573]Again, this is the trends in microhardness
[00:09:50.489]compared to the trends in NCSE;
[00:09:53.490]There were less obvious conclusions from this one,
[00:09:59.707]but again the hybrid does consume more energy
[00:10:03.090]than the as-printed, other than in a couple spots,
[00:10:06.409]but this 5.5 and 6 mm distance from the surface
[00:10:11.160]on the as-printed part on the bottom graph
[00:10:13.393]corresponds to where we can see the spike
[00:10:17.094]in the as-printed energy consumed color chart here,
[00:10:21.344]in the center of that red column,
[00:10:24.612]which is an encouraging data point
[00:10:27.396]because it reinforces our conclusion that microhardness
[00:10:31.693]and energy consumed during when milling a part
[00:10:35.128]are connected.
[00:10:37.093]Here are our conclusions;
[00:10:38.476]the hybrid sample consumed more energy overall;
[00:10:41.970]the energy consumption tended to increase per layer
[00:10:44.405]from top to bottom in hybrid and wrought samples
[00:10:46.723]but peaked in the middle of the build, or 5 to 5.5 millimeters
[00:10:52.157]from the surface in the as-printed sample,
[00:10:55.173]and the microhardness, as I mentioned,
[00:10:57.081]does show a loose correlation to the NCSE.
[00:10:58.908]And here are our resources.
[00:11:02.575]Thank you!

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

UCARE Summer Symposium Presentation

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments