SFSA T&O Conference 2017

Size: px

Start display at page:

Download "SFSA T&O Conference 2017"

Brianna Cobb
5 years ago
Views:

1 Air Entrainment: testing the model. Joe Plunger, President/Owner Midwest Metal Products, Inc (MMP) Acknowledgements and thanks for their considerable contributions to this work: Dave Poweleit, Christoph Beckermann, Seyyed Hojjat Majidi, Lucas Archer, Jerry Theil, Travis Frush, Dave Michalski, Brian Maloney, Paul Bambenek, Jeff Thorn, Kurt Thorn, plus other members of MMP s crew, University of Iowa, University of Northern Iowa. joe@midwestmetalproducts.com SFSA T&O Conference 2017 The 2016 presentation on modeling of air entrainment (Item 4.8 from 2016 T&O documents) done by Seyyed Hojjat Majidi and Christoph Beckermann in combination with events of earlier that year peaked my interest in learning more about how a naturally pressurized gating would test with that program, that developed into pouring samples for evaluation against the models. Summary: Modeling of Naturally Pressurized vs other variations on gating showed the program would predict less entrainment, which should lead to less defects related to air in parts made with that approach (see Fig.4). 4 variations of gating were modeled with the MMP modified Puhakka Top Gate design coming in with best result. We moved on to test the 4 variations, pouring 2 separate heats at MMP. Conclusion from this two heat test: Results were not dramatic, nor did they clearly show significant difference in the approaches used for the test. There were execution problems in the actual testing that may ultimately be shading the results. (see notes in Lucas Archer s presentation included in this paper) I continue implementation at MMP bolstered by excellent results on job starts across different alloys and configurations, some examples are at the end of this paper. 1 P age

2 Contents: Page 3. Background Page T&O Page 5. Idea leading to this study Page 9. Test information Page 13 Lucas Archer report on test results Page 22.MMP part examples 2 P age

3 Background: Earlier in 2016 I had actively followed and participated in the AFS forum discussions regarding John Campbell s work and Bob Puhakka s implementation of principles of naturally pressurized gating. I followed the discussions and did all of the research I could via the internet along with buying and reading John s 2015 edition of the Complete Casting Handbook. As a result I became convinced that this approach had merit, I had drunk the Kool Aid so to speak. As a result of that interest I had a gating method done by Puhakka on what was a chronic bad actor stainless valve. MMP poured our last few replacement parts of what had been a long running and painful cycle on the only order we made on the part that year with the revised gating approach. There was considerable skepticism internally by my well-seasoned staff on the approach but we made the changes and successfully poured the parts which subsequently were successfully machined. This had all taken place prior to the T&O conference of last year. I was at that point continuing to study the published information to try and better understand the math. 3 P age

2016 T&O Many of us will recall the couple of sessions on the application of the principals involved in John Campbell s work a colorful admonishment might be the best description of Bob Puhakka s

8 session was presented on the Air Entrainment modeling, looking at shapes of sprues, pour cups, metal stream location, metal introduction level, and ultimately what effect a pouring basin had on

4 2016 T&O Many of us will recall the couple of sessions on the application of the principals involved in John Campbell s work a colorful admonishment might be the best description of Bob Puhakka s two presentations with the goal he expressed to me of shaking up the status quo and getting us foundry folks to look at our applications of metal introduction to molds. When the 4.8 session was presented on the Air Entrainment modeling, looking at shapes of sprues, pour cups, metal stream location, metal introduction level, and ultimately what effect a pouring basin had on the factor of air entrained during a pour sequence, it resonated with me as providing additional support for the work of Campbell and Puhakka that I had already been working on. Figure 1 Beckermann & Majidi 2016 T&O 4 P age

The idea: As a result of all of the above I contacted Christoph and Dave P after the conference to inquire on running the Top Gate method gating through the program to see what effect it had.

Figure 2 shows old and new geometries. Figure 2. MMP original gating vs Naturally Pressurized Top Gate.

5 The idea: As a result of all of the above I contacted Christoph and Dave P after the conference to inquire on running the Top Gate method gating through the program to see what effect it had. Christoph agreed it could be done and asked that we also re-create the geometry of our former gating so that they might be compared for the effect of the approach. Figure 2 shows old and new geometries. Figure 2. MMP original gating vs Naturally Pressurized Top Gate. ( Pour basin shown full but drains to dam height in practice and sprue to riser height ) David Michalski of MMP worked with Hojjat Majidi on the geometries that would be used in the program runs for what was the original idea of just looking at how the program would evaluate the two gating systems. Initially the old and new systems were compared and figure 3 shows the summary of that evaluation indicating that the new system would improve the air entrainment by a considerable amount, this work was done in February P age

6 Figure 3 From Majidi & Beckermann run on MMP geometry for valve gating. The two versions of the new gating in figure 2 were driven by variations in filling flow rate parameters to get close to MMP actual timed fills of the molds made. In either case it shows significant improvement in the air entrainment with the new gating. In July Hojjat provided a report with 2 additional gating variations run through the simulation. One designed by the University of Iowa and the other by the University of Northern Iowa. It was at this point that Dave P. suggested that we run some actual pours of each of the 4 versions to test the theory. The UI gating was MMP s original gating but with a pour basin added, the UNI was similar to the Puhakka Top Gate but with a different runner path and slimmed down up-gates. 6 P age

7 All of the pour basin versions are significantly better than original. Figure 4: From Seyyed Hojjat Majidi report of July 28, 2017 A point of order on the gating naming convention and version we have been implementing at MMP. The version MMP is primarily using is Naturally Pressurized that is, with the sprue cross section made to contain the natural taper of a falling stream of liquid. It is a variation that Bob Puhakka created and named the approach Top Gate. This is NOT the Vortex gate but instead employs tangential surge overflow cylinders on the gating beyond Top-gate configurations, these shock absorbers provide for a gradual slowing of the metal in the runner as backpressure builds. The Top-Gates complete the process with metal passing through a filter up a cylinder and into the part, and in so doing we achieve the final slow down to the desired velocity of fill to avoid the mixing up of the oxide film (Bifilm) on the top of the rising bath. The Vortex gate incorporated the shock absorber as a racetrack of the gating around the gate cylinder, see Figure 5, however this was found by Puhakka in testing/application to still introduce too much velocity/mixing of the Bifilm in areas that would allow some to enter the cavity and has since been abandoned by him. 7 P age

8 Figure 5. MMP has not done any work with Trident style gating, see Campbell s 2015 book for detail, we were advised that it would be overkill for our part size and metal drop heights leading to unnecessary expense. 8 P age

MMP poured them in two separate heats in CF8M stainless along with 8 of the test plates using geometry provided by UAB which matches geometry

9 The tests: Through the efforts of everyone mentioned in the leading acknowledgements we were able to have all 4 versions of the gating made up in printed molds from UNI. MMP poured them in two separate heats in CF8M stainless along with 8 of the test plates using geometry provided by UAB which matches geometry used in some past studies to provide continuity for additional evaluation. The two heats were poured at MMP one August 23 rd and the other September 27 th. Assorted photos of the processes follow: Molds as received at MMP Kurt and Jeff closing up Top Gate design. 9 P age

10 UI design set up to pour Heat 1 UNI design post pouring Heat 2 MMP Original gating tree. Top Gate tree (bottom view). 10 P age

11 UNI gating tree (bottom view) UI gating tree. Top Gate pouring Heat 1 Stopper Stuck and splash guard rolled into stream causing an interruption in the pour problems were observed in evaluation. 11 P age

12 Screen shots from pouring video of UI mold in Heat 1, pour basin used with no stopper. Lucas set up and evaluating the castings and poured plates at MMP. 12 P age

13 #&!*C(15('#&-!'.*!;(,'#&-,!0/&*!/C*$!'9/!C#,#',!'/!::4!(&0!(11!'.*!#)(-*,! "#$"

14 !&"#"! "#$"!

15 !'"#"! "#$"!

16 !("#"! "#$"

17 !)"#"! "#$"

18 !*"#"! "#$"

19 !+"#"! "#$"

20 $,"#"! "#$"

21 $!"#"! "#$"

950-2100 lb pour wt, part weights ranging

22 MMP success stories implementing the Naturally Pressurized approach with Up-gates and Shock Absorber cylinders. Series of 3 die blocks in 440B Stainless lb pour wt, part weights ranging from lbs. All made with similar approach. All passed level 2 x-ray and machined up without flaws. NO welding done on parts. 22 P age

source that regularly provided parts that

Previous gating required 4+ hours of initial

First run with improved gating required 30

23 Gear blanks in low alloy steel lb pour weight 941 lb casting made with similar approach. First pass NO welding done to part, machined up with out flaws - Client was estastic having suffered for several years with a source that regularly provided parts that required him to repair post machining. Explosion containment housing in 316L stainless, part is upgraded post x- ray. Previous gating required 4+ hours of initial repair and another 2+hours post x-ray. First run with improved gating required 30 minutes pre x-ray and 10 minutes post x-ray, some improvements were made following that run, yet to see full evaluation. 23 P age

The Design of Gating System 4. Design of gating system elements 1

MME 345 Lecture 17 The Design of Gating System 4. Design of gating system elements 1 Ref: [1] P. Beeley, Foundry Technology, Butterworth-Heinemann, 2001 [2] J. Campbell, Castings, Butterworth-Heinemann,