FLOW SIMULATION / COLD LAP / GATING DESIGN

Hi.
Wanted to give some good example on what we do in foundry when failure happens time to time and there is need for more detailed investigation on matter and optimize casting gatings system to ensure good result at different sand, metal or other production parameters.

Cold run defects were seen from time to time in production of this ductile iron cover. It was first thought that this was due to an error in pouring or varying production parameters. I decided to investigate this closer in casting simulation.
First initial simulations were done to evaluate existing technology. It was seen in simulation that at low pouring speed the defect can appear in the exact place as seen on casings.
Gating calculation was performed as well to estimate correct filling time.



From the simulation it was seen that the cold lap is created somewhat because the metal gets too early in the inner circle where the casting is only 4mm thick and the metal is not enough to fill it fast enough.

This is why the solution to incline the gate and fill the outer circle first and later inner circle was simulated.

Here you can see 2 simulation with only varying gate direction side by side and synchronized by time. You can see that after 3 seconds in the design with straight gates the inner circle metal get colder and starts to solidify (darker red color) In the picture on the right it just fills the outer ring.


Here is video of the 2 comparison:



After this change it gave these benefits:
* Reduced scrap because of cold run to 0
* Improved the filling of the cavity even at lower temperatures and slower pouring times

It just one example to show how gate direction and place to cast geometry can change the final result.

 

That is an interesting simulation. Thank you!

In my tiny foundry with no simulation software other than my imagination I would incline flat molds like this plate so that the runner was maybe 1/2" an inch lower than the furthest portion of the mold cavity. That is because, when the mold is level, I would expect the sort of random sloshing of metal seen at around the 20 second mark. Were the mold inclined, I would expect the filling to be more organized beginning at the left and progressively filling to the right with both the inner and outer ring progressively and steadily to at least some approximation. Would it be possible to run a simulation with the mold inclined just a few degrees?

Denis

Added: Heck, if I am going to ask favors, why not go for it? What would be really interesting would be (providing it does not involve a lot of computer time or work) to run a sim at, say, 1, 3, and 5 degrees to see just what differences there are and maybe get an idea of the minimum inclination to have a substantial effect.

 

The old style gating had the ingates at 90 degrees and had to build up pressure to push the metal through the ingate as inertia pushed it past the ingate opening. The new design is flowing in almost the same direction as the runner and inertia pushes it into the ingate causing hotter iron to get in the casting sooner. Im with Kelly on being able to recognize some metal savings on sprue size...seems pretty large. It looks like you are using a filter and probably using a sprue that will utilize the entire area of the filter. We used to do that too, when we ran filters, but we would taper the sprue normally and then open the bottom of the sprue up on the very top of the filter to utilize all of the surface of the filter. Depending on how tall the sprue is and how many molds are produced daily, that could be significant metal savings. Also having dealt with bad pouring and bad pouring habits the cost savings may be justified. Sometimes those bigger sprue can save you if you have a bad pour.

 

Yes. This is what we always look for - worst case scenario. It can be that pourer do not keep the pouring basing full and for such situation the design have to ensure that castings still turn out good. We also investigate feeding of any casting for worst case scenario when designing the feeders. Software allows me to change the filling speed, angle, height. And I can change the pouring speed at specific time after start simulation real pour (for example usually human start faster and then slows down when the basin gets full).

Here is pic of the bad castings we got in past.

Yes you are right it suggests that sprue can be smaller. This is something that we have choose to keep standard sprues for all castings. Also the pouring basin is drilled on automatic molding line and are the same for each mold regardless of casting.

 

I did made simulation by tilting the whole by 3 deg. Sorry will take too much time to make all the 3 simulations. As you can see it does make smaller splashing. But i am not sure how accurate this simulates splashing as it might be related to the size of the elements. But the tilting seems to make a bit worse as the metal slowly gets in thin part and start to cool down. Have to say the parameters on 2 simulations are not 100% same, but hope it gives some idea.

Anyhow, tilting is not possible for us in foundry anyway , but sure in some applications castings are done by tilting.

 

That is an interesting simulation, for sure. I did step through it slowly especially around the 25 second area and beyond. Thank You!!



It does appear that the center ring stays hotter and more uniformly so in the inclined pour in the right-hand simulation.

The piece above is very similar to pouring a plaque as it is pretty thin and relatively flat. Many many times we have had folks (myself included) pouring plaque-like patterns and having problems with cold shuts or laps. I can not remember an exception to the problem being solved by slightly inclining the mold---maybe 2 degrees or so.

In my small (65-70 pound max capacity) iron foundry I have switched to inclining virtually all of my molds as that has eliminated problems with cold laps. For instance I had been casting the following with the mold level.






Yes, my pouring basin is overly large on this old pour as it gave me bigger target to hit with my single-handed pouring trolley.


I was getting good results most of the time with the configuration above. But every fourth pour there would be a maybe 3" subtle crease on the interior at one end or the other. This was not a fatal flaw but it still was not ideal. I attributed the cold lap to the fact that metal could flow either right or left depending on least resistance to flow. So, flow probably randomly increased or decreased left or right throughout the pour.

Once I regated and poured inclined, this problem never recurred. The same is true on a very similar piece that is 26" long rather than 18 as is the case with the above pattern. In all cases the iron temp was 2550 as poured. There are other examples I could cite, but the above should convey the idea.

Too bad inclining the mold is impossible as I really think it would be likely to help.

Denis

 

I dont think I have ever seen a mold tilted in a high volume production environment. I have seen some other things but not tilted. The nice thing about hobby level... there is basically no restrictions and you can virtually do whatever you need to do to make your life easier. In the production world we have to design systems that do not fail with the set conditions we have. That being said we would have to change the gating system to eliminate the misrun, shortpours and cold lap defects out of the casting. I have a few jobs here I would tilt the shit out of, if I could.LOL

 

But, Billy, it is hard to imagine a mold any more tipped than the ones poured by Lodge in this video!



I mean these pans are vertical. I suppose you must have other stuff that is cast horizontally. If I am ever in your area, I am somehow going to cadge a tour.

Denis

 

The scale of things at Lodge is amazing.
Thanks everyone for the simulations, videos, and discussion.

 

We used to have many horizontal molding lines but they were no match for the speed and consistency of the Disa vertical molding machines. Disa recommend using up to 65% of the plate for casting and gating so we stay at around 80 to 85 to insure we are being as efficient as possible.LOL
I can give tours but have to get them approved in advance. They also do tours during the National Cornbread Festival for everyone who wants one. I think you would rather have the plant in operation so you can actually see the whole process.

 

https://nationalcornbread.com/

A whole festival for cornbread, who would have thought?

 

My ten-year-old grandson was over the other night and foolishly asked me how things were going at the foundry. The above discussion was fresh in my head and I launched into telling him about regating some of my molds and why. He courteously said "That's brilliant, Grandpa." (He knows how to butter up Grandpa.)

To try to illustrate what I was talking about, I pulled out a glass 8X10 casserole dish (aluminum sheet pans are usually too warped) and put it in the sink level and let water run into it from one end at a modest rate. As the water started to fill the dish a confused and wandering fluid front meandered across the dish. So, then I dried it out and inclined the dish just a quarter inch or so and again repeated the filling at about the same rate. It was obvious to him how smooth and steady the filling was. I was surprised at how graphic the difference was. I summed it up for him by suggesting it is best to not give the molten iron chaotic choices about which direction to flow but rather force it into a steady smooth stream.

We also looked carefully at the simulations provided above. Given the fact that he is growing up in a video world, he loved the sims too. One thing we noticed that I had not grasped before was that in the inclined simulated pour the sprue is also inclined and the metal can be seen to be spiraling down the sprue rather than just flowing linearly as it does on the vertical pour.

That whole interaction lasted maybe ten minutes, But it was cool.

Denis

 

That's really cool! I show my daughter stuff all the time but most of it she has no idea what she is looking at.LOL

 

If I missed the answer to this question I apologize. What software are you using? Excellent work.

 

I work with NovaCast

 

Brilliant thank you, will check it out.