CASTING SHRINKAGE BEHAVIOUR ANALYSIS 5

[skengg]

Hello all,
i just wanna know some deeper stuffs on how the casting shrinkage in solidification stage works and what are the factors which affecting it.
like we design sand casting pattern/dies and for steel we use 1.5% shrinkage as per industry norm on pattern. but while inspecting we find that some of the dimensions on the part shrink according to 1.5%, while some of the dimensions shrink more than expected. so I want to know how this thing work, currently on 3d model we just use uniform scale command to increase the pattern uniformly 1.5% from every direction.
can any one share insight view or references to study more about it.

 

[mfgenggear]

Skengg
Inspect and record, for historical data and adjust.
No silver bullet to theoretically solve this.
Hopefully the mold can be modified. Lost wax pr ocess . Also notify supplier and work with them.

 

[EdStainless]

A portion of this is the overall change in the part size and shape.
Variations in section thickness play a lot of role in the size and shape changes.
I would suggest that you not focus on linear dimensions by try to map the full 3d shape of the part.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[mfgenggear]

In addition would recommend inspecting critical dimensions were there will be post machining. And it had to clean up. It's better to be on a stock on condition. It can be taken off but it can't be put back on.

 

[dvd]

Just comments from the cheap seats, I have no expertise in this area. I can picture that the thicker sections will still be solidifying when thin sections are solid, but these solid thin sections do not have their room temperature material properties (i.e., they have different high temperature young's modulus, poisson's ratio tensile yield strength, compressive yield, etc) so as the thick section solidifies and shrinks it distorts the thinner, solid, but hot and weak sections.

 

[3DDave]

Another factor is that there are two main sources of shrinkage. One is simple thermal coefficient of expansion. The other is the way the crystals are forming. Because cooling is rarely at exactly the same rate throughout the casting the formation of crystals vary throughout the casting. One needs a soak at high temperature followed by slow cooling to cause crystals to be more uniform through the material in order that only thermal coefficient of expansion is in play.

There can also be distortions from non-uniform cooling as the frozen metal that has cooled applies pressure to the still plastic areas of the material; this is the way that tempered glass is purposely created, as is the curve of the katana blades.

Too add - this is why parts can distort when heat treated. There are built in stresses from the original non-uniform temperature transition and the crystals can change form - particularly the BCC vs. FCC in iron-carbon metals.

 

[mfgenggear]

Yes agreed, there are many factors.
The chemistry of material. Temperatures.
The geometry of the part, the size.
Thin section will cool faster, thus causing
Distortion. Transformation of the grain structure. Ect

 

[skengg]

thanks all for providing their insight views, i know that there are lots of factors which influence shrinkage behavior of casting, but i want to know how u approach to make the pattern/die of any part, like i have casting part of which i need to make sand casting pattern, i usually make 3d models of casting part and add uniform scaling to compensate material shrinkage allowance (for example steel casting which shrink around 1.5%), i made the pattern 1.015 times larger than cast part uniformly. even though for all the critical dimensions we provided the machining stocks on every face, so at the end we achieve our dimensional requirements. but still whenever i try to inspect the as cast part, its total length or other dimensions not shrink as per expectation. our company rarely do casting dimensional inspection until and unless any things get wrong, only visual inspection prefer.
also, i wanna know this behavior because in case of investment casting we make die and most dimensions of any part usually stay as cast, so achieving dimensional requirements in that as cast condition become very tough.

 

[3DDave]

The castings I worked with - the supplier would take a shot at estimating shrinkage, fully inspect the finished casting, then take any unwanted differences to the pattern maker to alter the pattern to compensate. On larger castings they would work with "chills", slugs of metal to cause the nearby metal to cool more rapidly and make thicker sections cool at a rate similar to thin sections.

It may be there is a hyperphysics FEA that can predict the conversion from liquid to crystals and manage the specific changes in energy content and thermal conductivity.

 

[Tmoose]

I've read that the stiffness/strength of the mold will directly affect how the solidifying metal can or must shrink.
The resulting issues especially with a sand mold can be changing dimensions (weak mold?) or even cracking ( strong and stiff mold?).

Hence the cut and try development mentioned by 3DDave.

For extra fun the wear liners we make are 10-20 inch irregular shapes, with thickness from .5" - 2".
Usually one of the nasty ASTM A532 irons. They seem to go thru multiple changes when solid and cooling.
Depending how the foundry lets them cool the different properties at different temperatures mentioned by DVD and 3DDave can and will conspire to make a casting that had every reason to be flat curl up .25 " or more.
A machining allowance of ~.125" that worked okay from the 1960s all of a sudden is not enough for parts made in molds created using the same pattern !

 

[EdStainless]

We used to cast high Cr white cast iron cylinders.
They had to come out the correct size because we didn't want to take days to grind each one.
The center cores were hollow so that they would collapse as the metal cooled.
And we used a large pad for feed with 5 gates into the part.
If a mold didn't get uniformly compacted, you would end up with parts that weren't round.
And you couldn't change sand, it had to be kept identical down to the brand of each additive.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[CWB1]

My employers have had mold flow packages so I invert the model to form the cope and drag, add gating and risers, and run the sim.