We have already published a case study of the CFD simulation of propagation of airborne pollutants, in which we investigated spreading of carbon dioxide in a chemical plant. Emitting CO2 was part of the technology and happened 50m above ground level, but during the operation of industrial plants, unwanted substances are sometimes released much closer to the operators handling the machines.
This includes propagation of dust and smoke generated in many ways, but the leakage of harmful gases from lab equipment can also be considered. Computer simulation of these processes is especially important when large amount of pollutants are formed in a closed space, for example, in one of the production halls of a cylinder head foundry that produces 3 million castings a year.
Audi S8, R8, S6, RS6, Lamborghini Gallardo, Huracán. What is in common with these cars? It is the fact that these were all built with VW’s V10 petrol engine and it is highly possible that the cylinder heads were made in one of the largest foundry in Central Europe: in Hungary. Not all of the 3 million are for V10 engines, the majority of the production are made of 1.5L or 1.6L diesel or 1.4L, 1.6L or 2.0L turbocharged petrol engine cylinder heads.
Let it be any size and type, one thing is sure. If a foundry uses a casting technology like the one in Hungary, then the factory must calculate with the significant and varying amount of smoke generated in the process.
But what is this smoke anyway? Where does it come from? When, and most importantly how much is generated?
Well, this smoke comes from sand cores forming cylinder head cavities like water jacket, intake and exhaust channels. In our case sand cores are made using organic resin, which burns with thick smoke when 700°C molten aluminium is poured into the die.
As temperature of the aluminium decreases, sand cores get hotter and hotter and thus more smoke is generated. The amount of smoke increases suddenly when the die opens and smoke can finally escape through all surfaces of the cores into the atmosphere.
Smoke generation is part of the technology here as well. On the other hand the challenge for the foundry was that productivity had been increased so much in the last decade that smoke extraction system of the casting machines were unable to cope with the increasing amount of smoke generated by the increasing amount of castings. Extraction technology had to be developed to match a production volume ever larger than the current one.
However, the cleaning capacity of the filter is finite and the company did not plan to sink a couple of millions into a new filter. So the existing filtering capacity had to be redistributed, more efficiently used. Here comes CFD simulation in the picture.
We got the job of redesigning covers and extraction systems of the casting machine by using CFD simulations. The set aim was that – in an optimum situation – no smoke should escape from the casting machine. At the top of all that we had to make it so that extraction volume flow was limited to 8500 m3/h per machine at 60°C air temperature.
When working on a project this magnitude, the first step is to determine the initial status of the machine’s smoke extraction capacity and to reproduce it in a simulation. This way we have a set of boundary conditions in the simulation and a good quality CAD geometry which provide the foundation for creating new extraction configurations. Also we have a common ground to compare new versions to the original.
The casting machine producing the heaviest cylinder head was chosen as a test equipment. We build its 3D CAD geometry and shot several videos of full casting cycles. The picture below shows the casting machine from the front and size of the operator in black well represents the machine’s dimensions.
The steel sheet metal structure of the casting machine’s extraction cover was made transparent in the picture so that the two orange coloured dies with castings remained visible. In the simulation model we created four important positions where castings emit smoke during the casting cycle. Three out of four are in the operator’s level of work, the fourth is in the brown box of the automatic manipulator 2m above ground level. The manipulator is equipped with smoke extraction as well.
Red parts show areas in which from time to time auxiliary machinery are at present, so these were the areas we had to avoid while thinking about modifications. On the left there was the automatic ladle on the right there was the manipulator which removes freshly cast cylinder heads from the die.
On the rear side of the casting machine – shown in the picture below – there were the blue coloured extraction pipes which were connected to rather small extraction hoods on top of the machine cover. There are three of those small hoods in the original design. The bright red pipe – slightly smaller than the other three – is the one that extracts smoke from the manipulator box.
Videos of the casting cycle recorded how much smoke escaped the extraction cover and where it entered the atmosphere. Videos also showed how long a casting cycle was and how long normal and high intensity smoke generation were at present.
The four numbered casting and smoke genereation positions were results of a rational simplification. In real life in this casting machine there are only two dies and two castings at present in one time. The dies however move synchronised in a horizontal plane and they switch places in front of the operator, who is standing in the middle, at the open space of position Nr. 3.
Though in SC/Tetra it is possible to make the dies move, we decided to have the dies standing still and make the location of smoke generation change only. The four numbered positions with four casting CAD models represent the stations where the moving casting would emit smoke most of the times in a casting cycle. Casting models placed in each of the four positions generate smoke in the simulation synchronised with the smoke generation cycle recorded in the videos.
Pictures below show the defining smoke generation positions of the casting cycle. Blue surfaces show generated smoke.
Red arrow in the above picture shows that in the next stage smoking casting of position 1. will move to position 3. after the die in position 3. containing only cores is filled with melt and moves to position 2. Movement of from position 3. to position 2. is shown by the blue arrow. The die freshly filled and moved into postion 2. will start smoking with normal intensity.
In the 2nd stage of the cycle extraction system has to handle two smoke generation positions as it is shown on the picture below.
In position 2. the casting is at the beginning of its solidification and emits smoke with normal intensity. On the other hand the casting that moved from position 1. to position 3. is at the end of its solidification, waiting for the die to open, sand cores are really hot and the amount of smoke emitted is steadily increasing. Die opens in the next stage which is shown in the picture below.
When die opens in position 3., intensity of smoke generation is suddenly 3 times more intensive than under normal conditions. The extraction box of the manipulator above position 3. is unable to handle such amount of smoke neither in the simulation nor in the reality. It is clearly visible in the picture above that a blue cloud of smoke escapes at all sides of the extraction box.
The intensively smoking casting will be removed from position 3. by the manipulator in the direction of the red arrow into position 4, though the extraction box stays where it is, not sinking with the manipulator arm. Simulated smoke generation ceases here because manipulator moves to the right side and places the casting to the next processing station.
The casting cycle continues and position 2. emits normal intensity smoke while operator places fresh cores into the empty die in position 3.
Red arrow in the next picture shows that in the coming stage of the cycle, the casting in position 2. smoking with normal intensity will slide to position 3. At the same time die in position 3. after receiveing molten aluminium will slide into position 1. (blue arrow) and start smoking with normal intensity.
In the final stage die in postion 3. opens, high intensity smoking happens and after loads of smoke escaped the extraction box of the manipulator, the whole casting cycle starts over again. Here is an animation of smoke generation in the whole cycle:
All the smoke these three extraction pipe connections on top surface of the cover and the extration box of the manipulator are not able to collect, goes into the workshop air.
This is a rather unpleasant situation. The first problem is that smoke from positions 1. and 2. escapes under the cover at the height of the operator’s face. The second is that there are several other machines with a cover like this and their emitted smoke fills the workshop air quite fast. This leads to the requirement that all of the smoke must be kept within the casting machine cover and the extraction system connected to the cover must be improved so that it is capable of collecting all the smoke of an even faster pace of casting production.
The next part of the case study will be about how we did it and how the new casting machine cover looks like. Stay tuned.