Paul Withey, Casting Specialist for Rolls Royce
Chris - This is the Naked Scientists. I'm Chris Smith. If you've been on a flight recently, then there's a very high likelihood that you were carried aloft by a jet engine. These work by compressing air and squirting in fuel which then burns, expands, and generates thrust. But this also creates one of the harshest imaginable environments inside the engine where the gas stream routinely exceeds 1500 degrees Celsius – that's well beyond the melting point of standard metals. To withstand this takes specialist superalloy components which are produced with a very high precision and to find out how they're being made, Ben Valsler has been to the casting facility of one of the world’s top jet engine manufactures, Rolls Royce in Derby, where he met Paul Withey.
Paul - I'm the casting specialist for Rolls Royce and this facility is the precision casting facility in Derby. It’s Rolls Royce’s largest foundry and it makes turbine components for large civil engines and military engines. This is one of the single crystal facilities in the world which is one particular branch of precision casting that makes only single crystal components and the benefit from an aerospace perspective is that if you take out all the boundaries between each of the crystals, you get the strongest possible metal that you can, and that's really needed in the hot part of the engine. So like a sugar cube when you crush it, you actually don't break each of the individual grains, but you break the boundaries between the grains. It’s the same in aerospace metal, so we make single crystal components to get rid of those grain boundaries and have the strongest part we can.
Ben - So what's the actual process? How do you go from raw metal to jet engine parts?
Paul - There's a number of different stages. If you imagine the component as not being a solid component but being a hollow component, we have to make a ceramic model of the hole we want. So we make that component called a core, inject wax around it, so we end up with the wax model of the component with the hole in the component filled with ceramic. Then we take that model, we’d go into what we call the shell room where we actually cloak it or clothe it in ceramic, and then we’ll replace the wax with the metal and then there’ll be the solidification process that's the heart of this process that means we can grow a single crystal. We have to pour in a vacuum so everything is hidden away behind steel walls.
Ben - Could we have a look anyway?
Paul - Of course. Let’s go.
…This is what we called a wax room. It’s a room in which we take the initial ceramic cores that we talked about earlier and we’re going to inject the wax around them. The room is a mixture of wax injection machines where we’re actually putting the steel dyes in there and injecting the wax into the shape we want, and assembly benches where we’re taking those wax shapes and putting them into the orientation on the runner system that we need for the further processing of the parts.
Ben - So now, can we follow some of these wax structures through into the next stage?
Paul - Of course we can.
Ben - We’ve come through a large door and into the second stage. The last room was very warm. It smelled of melted wax. In here, it feels a lot colder, but it’s also a lot noisier. What happens in here?
Paul - Now this is a shell room. It’s actually no colder. It just feels colder because of the controlled humidity, but actually, it’s held at exactly the same temperature so the wax is exactly the same size in one room as it is in the next, and what we’re going to do is take those wax assemblies, we’re going to clean them, wash them, and then we’re going to dip them in a ceramic slurry, and then take them out, drain them, and then sprinkle grit on them, and then allow them to dry for a couple of hours and then bring them back and do the whole thing over and over again up to 10 times.
Ben - So, the slurry is sort of the sandy stuff that will turn into your ceramic layer.
Paul - It’s actually a liquid but actually has a bit of a ceramic glue in there called a silica sol that will allow that all to stick together and the bulk of it is delivered as a gritty dry sand sprinkled on top of the wet surface.
Ben - So how thick a layer of ceramic do you end up with?
Paul - It’s normally somewhere in the region of about 5 millimetres.
Ben - And where do we go to next?
Paul - The next stage of the process is to take the wax out and then move on to casting.
Ben - So the wax moulds are now invested with their layer of ceramic and we’ve come through into a very, very warm room. What happens in here?
Paul - This is the room where we take the wax out. We wheel the moulds in, we close the door, we press the button, and in 4 to 5 seconds, we inject superheated steam at 180 degrees which allows the wax to melt out really quickly. The wax doesn’t melt as a whole. It only melts on the surface and doesn’t conduct heat well. If we did it slowly, the wax will expand more than the shell could take and we’d blow the shell off the outside. So this way, we get the wax out without breaking the shell.
Ben - And now, we’re actually left with a ceramic mould that we can then pour metal into?
Paul - We now have a mould with a core in the middle of it held in space that's now ready to cast.
Ben - And now, with the ceramic layer fired, we actually get through to the bit that involves molten metal, the bit that most people think of when you talk about casting. It doesn’t look like I would imagine it would - I can't see these rivers of molten metal.
Paul - The whole point of this process is that we control it as much as we can, because we have to take that mould, we lock it onto a copper cooled chill at the bottom and that allows this to have a very cold end of the mould. We then push the whole mould up into a chamber that’s at around 1500 degrees C, allow it to bake there for about 20 minutes and then we melt a charge in a crucible above that, pour that charge into the mould itself. The mould at 1500 degrees C is above the melting point of the metal, so most of it stays molten except for the last little bit that hit the copper chill and then froze as lots of different grains, all orientations, hundreds and thousands of grains. We then take the mould and we draw it into a cold chamber so there's a temperature gradient from hot to cold and those grains all start to grow. It just so happens that the grains that grow the fastest are also the grains in the orientation we want for mechanical properties purposes. So by a natural selection process, they grow really quickly upwards and sideways, and kill off all the other grains.
Ben - So you're getting a single crystal because that one crystal structure happens to just be a bit more stable and perhaps outcompete the other crystal structures that could form?
Paul - That's correct.
Ben - So how long does it take for this single crystal structure to form?
Paul - It takes just over an hour to make a single crystal component of the size that we tend to manufacture and that’s for the large civil engines of a component that's about 10 centimetres to 15 centimetres tall.
Ben - So once they've been poured, once they've had a chance to set, where do we take them now?
Paul - They then move to the back of the facility where we’ll cut-off each of the individual components and then use a strong alkali solution to remove the core which then leaves us with these castings with the right hollow inside them in the right place.
Ben - So once the products have come through, they've cooled, they've had all the bits of excess cut off, how do you check that they are what they need to be?
Paul - We have to check that they are a single crystal, and we’ll do that by dipping them in acid and then looking at whether there's a grain structure there, and if there isn’t, we’ve got a single crystal. We’ll also check using x-rays to see whether they’re the right orientation of crystal, we’ll use gauges to check the external profile is correct, the length is correct, and that they're all at the right tolerances. We’ll also use a dye penetrant to look for surface defects and we’ll use x-rays to check for internal defects.
Ben - And how perfect do they need to be? What error margins do you have?
Paul - In terms of the crystal orientation, we’ve got a few degrees of orientation that we need to hit. In terms of dimensional requirements, the drawing usually requires us to be within 0.1 or 0.2 of a millimetre of the required position, and so, we’ll check every blade to those kind of dimensions. And unlike certain industries where you can sample inspect, once we’ve got a metal part, every part goes through every inspectional operation.
Paul - This is one of the finished blades and this is actually gone through the next stage of the operation which is machining and the drilling in of the film cooling holes into the internal core passage. It’s about 12 centimetres tall and it survives the really high temperatures and pressures and stresses inside the engine.
Ben - What conditions could that piece of metal now put up with?
Paul - Well this turbine blade itself will face conditions of the gas stream around it when the engine is operating being 250 degrees above the melting point of the metal and the stresses when it’s spinning around at about 10,000 revs per minute is the equivalent of hanging a lorry off the end of it in a static case. This blade then has to take 750 horsepower or more out of the hot gas stream to power the upfront parts of the engine. So, each one is about the same horsepower as a formula one car. That gives this a component that will last more than 5 million flying miles and it’s quite an arduous environment to live in.
Chris - Paul Withey, taking Ben Valsler around Rolls Royce’s precision casting facility in Derby. Rolls Royce reckon that at least 200,000 people around the world are being held safely aloft by their engines at any moment in time. In other words, right now.
do they rust considering they are made of steel steve frost, Tue, 17th Mar 2015
They said the blades are made of superalloy, which is basically nickel, cobalt and chromium (plus other stuff) but no iron, so they don't 'rust'. That said, superalloys at service temperatures do oxidize (burn) and therefore the trick is to make them oxidize slowly enough to be useful for 5 million miles or so. This is done with coatings that create oxide scales to seal the surface. Doug, Thu, 7th May 2015