Casting Basics

Now, let’s actually get into the process of making parts.  First we’ll go through casting.  Casting has been around for literally centuries – almost 6000 years by some accounts, because the process is so simple and intuitive.  It is the process of heating up metal until it becomes molten (liquid), and then pouring this into a mold of your desired shape.  The metal cools and solidifies into the shape of the mold – and now you have a solid metal component in that shape.  Historically, the metals cast were soft metals, since they having lower melting points, and with primitive methods it’s difficult to heat up metals past a certain point.  So silver and gold was used extensively, mostly for making coins and jewellry.  The casting of iron didn’t appear until many years later, sometime around 1200 A.D.  Nowadays, iron is still the most commonly cast material – you may have heard of the term cast iron.  This is a little misleading, because cast iron is a material, in the iron-carbon alloys.  It has a carbon content of greater than 2%, and some silicon as well.  In other words, just because a iron material has been cast, does not make it cast iron!  Cast iron flows well (‘good fluidity’) and has a relatively low melting point, which makes it excellent for casting – you can easily cast parts that are thin or have intricate features.  Furthermore, it is strong, but easily machined.  Historically, it has been used extensively in a range of engineering applications, although its use in some higher end engineering applications, such as automotive components, is declining.  It is still used extensively to make engine blocks of vehicles.  A major drawback is that it tends to be very brittle.

General Advantages of Casting

Let’s get back to a more general discussion about casting before we get into the materials that are used.  What are the general advantages of casting?  One of the big ones is that you can make complex geometries and detailed features relatively easily, compared to some of the other manufacturing methods.  Moreover, usually these complex parts can be cast as a single piece – so you can avoid machining operations or joining multiple pieces together.  You can cast many, many pieces (high volume) and you can cast massive parts that you probably couldn’t make any other way.  

Sand Casting

Let’s talk about sand casting; the mold is made of sand.  Sand.  I know, it seems a bit odd to be making automotive parts with a sand mold (engine blocks are usually sand cast).  The first step in making a casting is to make the mold.  Usually, the mold is two pieces that come together, with a hollow cavity taking on the shape of the part you want to make.  Then, you can add what are called cores – cores allow you to add detail to the casting – think of pieces that you add to the hollow cavity.  But let’s take a step back.  The very first step is actually pattern making.  Basically, you have to make a pattern, which makes the mold, and then you use the mold to create the cast part.  The pattern is filled with the sand, basically creating the shape of the sand mold.  Usually these are made out of wood, or metal or some sort of plastic. There are two halves to the pattern, called the cope and drag patterns, which come together.  Keep in mind that the pattern and the subsequent sand mold aren’t the exact same shape as the desired part, for a couple of main reasons: 1) there will be some shrinkage as the metal cools and solidifies, so the pattern/mold must be slightly larger.  This will depend on the type of material used, because different materials will shrink different amounts.  Additionally, it may be slightly larger in some areas so that material can be machined away afterwards – this is sometimes required because, as you can imagine, a sand mold doesn’t leave the best surface finish.  If you want to sand cast your part, but want a smooth surface in some areas, machining is definitely an option – basically, you machine away the rough surface.  

There are number of steps that are taken to create the mold.  First, you place the pattern basically inside of what is essentially a container (it has walls) and it’s called the flask.  Now, fill this container with sand so that it covers the pattern, and then compress the sand so that it closely matches the pattern as closely as possible.  Now you can remove the pattern, and you have half of the mold – this side is called the cope.  Do pretty much exactly the same process again – now you have the drag half.  The cope and the drag, the two halfs of the mold, both made of sand, which combined create the desired shape of the part.  Before you clamp them together, you can placed cores inside of the mold, to create cavities (intentional) inside of the casting.  The other term for them is negative forms.  Ideally, you can avoid the use of cores, because they add complexity to the casting and therefore cost.  But sometimes they can’t be avoided.  Next, you fill the mold cavity with molten metal, and let it solidify.  And then you can just break the sand mold away from the casting and there you go – your part in the desired shape.

Keep in mind that when you create the mold, there is space left to pore in the molten metal (called the downsprue).  There’s also features built into the mold, such as the riser, also known as the feeder.  The riser is basically a reservoir built into the mold, to prevent cavities due to shrinkage.  Basically, molten metal sits in this reservoir and continues to feed the mold with molten metal as required.  This works well to prevent cavities, but you need to make sure a few conditions are met first: obviously, the riser has to remain molten while the casting solidifies, otherwise it wouldn’t be much use!  The riser also has to have enough material to completely fill the shrinkage, and finally, the casting has to solidify moving towards the riser.  In other words, the portion of the casting closest to the riser has to solidify last, otherwise the riser won’t be able to supply material to the casting.  In terms of surface tolerance and roughness, well, it’s sand – but also it might be better than you expect.  Usually, the tolerance is somewhere in the range of 0.5 to 5 mm.  Surface finishes of somewhere in the neighbourhood of 50 micrometers can be expected, sometimes less.  What is perhaps most surprising in all of this is how quickly sand cast parts can be made, because it sounds like an archaic process.  In reality, hundreds of parts an hour can be made like this – with the help of automation.  Sand casting is relatively cheap, and extremely common.

Investment Casting, or Lost Wax Method

But what if we want better tolerance and better surface quality but it still makes sense to cast the part?  What if we want to avoid machining?  Although sand casting is common, there are other ways to make the part – it can be investment cast.  Investment casting involves making the mold out of wax – as with with sand casting, the mold is expendable, meaning that it’s not reusable.  Sometimes it’s called the lost wax process, because you literally lose the wax mold.  The basic process is as follows.  The first part is to make an accurate die of the mold, which can be used to make the wax patterns.  The wax patterns are usually attached into a tree form – with the sprue running down the middle and the wax patterns attached like branches, because this means that many parts can be made at once.  Then you take what is essentially this tree made of wax and dip it into a slurry (basically a fluid mixture of pulverized solid – in this case mostly silica, water, and some binding agents).  This is done multiple times so that the wax tree is evenly coated.  And once it is evenly coated, the whole thing is sprinkled with some silica sand and dried and allowed to harden, so that it’s basically covered in this stucco-like material.  Now you have this wax tree, with the trunk as the sprue (where the molten metal will eventually run down) and the branches the individual parts, and it’s covered and dried in this stucco stuff.  Bear with me here, we’re almost done.  The whole thing is heated so that the wax melts and runs out, and voila – we have created the mold.  Now all we have to do is pour the molten metal down the sprue, and it will fill up the individual parts.  Let it solidify, and then break away the mold.  And again, robots are used extensively in all parts of the process.

Die Casting

But we haven’t even covered the casting that you’ve probably likely heard about, which is die casting – like ‘die cast’ toys, those little metal cars that maybe you played with as a child, were likely die cast.  The basic premise is this: let’s not create an expendable mold, but instead metal is forced into a permanent mold (meaning that it is reusable) under high pressure – it’s not poured in like it is for sand and investment cast parts.  The molds have to be pretty strong and durable, like tool grade steel.  The cavity itself is created by bringing two mold halves together – so that they can be moved apart once the casting is completed.  One of the halves is typically fixed in place (the cover die half), while the other half can be moved (the ejector half of the mold).  Once the casting is completed, the component remains in the moving ejector half of the mold when the mold is opened.  The casting is then removed from the mold with high powered pins that push against the component (you can’t just tug it out, since it’ll be pretty snug inside the metal mold).  This is why the mold is designed so that the component remains in the ejector half, since this is where the ejector pins are located.  The ejector pins are designed with care, so that they are accurate and act with uniform force, so that the part isn’t damaged from being hit with one particular overzealous pin.  You have to consider the shape of the part, and incorporate draft angles – which basically mean that the solidified component can be more easily slid out.  If the part is wedged in tight due to the shape, then it might not be possible to remove the component!  So care needs to be taken in the design (again, design for manufacturing rears it’s head).  This whole process can be pretty rough on the molds, and so to extend the life of the molds (which are expensive), the mold can be sprayed first.  As well, the mold may have a cooling system to reduce the time spent waiting between making parts.  

A major drawback of die-casting is that the equipment and dies are quite expensive, and so it really only makes sense economically if large numbers of parts are being made (high production volume).  It is most suitable for parts that are small or medium size, and usually made of nonferrous alloys with zinc, aluminum, copper, and tin.  You’ll notice that steel isn’t on that list.  Many grades of steel cannot be die cast.  It’s just too difficult using this method, and would wear out the die at very high rates, but there are some very good options instead.  Aluminum alloys offers dimensional stability for complex shapes, and is lightweight, with good properties.  Zinc alloys are generally the easiest to cast, and it’s very ductile.  Copper is hard, with great mechanical properties – strength is closest to steel of all the mentioned alloys.   Although it depends on the metal, generally the maximum weight of the parts cast range from 10 to 70 pounds.  The dies typically last anywhere from 100k to 1000k (one million) cycles.  The main advantages of die casting are the excellent tolerances produced, the very smooth surface finish, the ability to produce very thin walls and intricate details, and the rapid rates of production.  What is particularly nice about die casting is that often you can avoid post-production machining, unlike sand casting and investment casting, which typically will require at least some machining.

Usually you’ll see two ways that die casting is classified: hot-chamber die casting, or cold chamber die casting.  I find the names a little misleading, since neither are really ‘cold’ – the metal is molten!  But as the name implies it refers to the chamber, not the molten metal.  With hot chamber die casting, a pool of molten metal feeds the die.  Molten metal is drawn into what as known as the ‘gooseneck’, and then the piston forces the molten metal out of the gooseneck and into the die.  The alternative is the cold chamber die casting, where the required amount of metal is first melted in a separate furnace, and then transferred to the injection cylinder, to be driven into the die by the piston.  The major difference between the two is cycle time – the hot-chamber casting method is quicker because metal can be drawn quickly out of the hot pool, whereas with the cold-chamber method the molten metal must first be transferred.  But the hot chamber system is difficult to use with metals that have a high metal point, because obviously the metal needs to remain molten.  This means that the hot chamber method is more likely to be used with zinc, in, and lead alloys, whereas aluminum alloys are primarily cold chamber die cast.

Design for Casting

Now that we’ve gone over the major casting processes, we can get back into the whole design for manufacturing discussion, because it really is an important subject for a engineer working in industry.  Whenever you’re designing a part, you pick a material and expect that you’ll get some specified mechanical properties.  With a cast part, getting good mechanical properties is a major concern – the mechanical properties can be severely degraded with a poor casting, making the part weaker and more likely to fail.  And so the casting process is crucial, and engineers must consider it when design parts to be cast.  One good rule of thumb is to have directional solidification – the part begins to solidify at one side and linearly solidifies to the other side, limiting cavities and defects which may form if it begins solidifying all over the place.  Sharp corners should generally be avoided, because defects can form there – and sharp corners are already a stress concentration, making that particular flaw a very dangerous one.  Fillets (the smoothing of sharp corners by adding radii) should be added wherever possible.  Some particular features of the part are also susceptible to local shrinkage during cooling, creating hotspots.  One of the major sources of this is changing thicknesses throughout the part, especially at junctions.  With different cooling rates due to the different thickness, you can end up with porosity.  It is generally a good rule of thumb to avoid adjacent thin and thick sections (i.e. rapidly changing thickness in a part).  This is unrelated to defects, but another guideline: keep in mind that it’s generally cheaper to add features to the casting that eliminate the need for post-machining operations.  For example, this could be a feature like threads, or lettering on a part.  Machining tends to be expensive.  The exception is when cores are required to make those features, because cores add complexity and slow the whole process down.  For instance, it is often cheaper to machine a hole rather than include it in the casting with the addition of a core.