In today’s fast changing world, research is going on to break every aspect of life into digits and finally to search the answers in scientific way. Few of the examples of this are: Recently Daily Mail1 published a news article saying that a doctor did research on the faces of celebrities and linked various ratios on their faces with Greek golden ratio Phi. After this study this doctor found that George Clooney is having most handsome face because it is very close to the golden ratio “Phi”. A list of celebrities based on the descending order of closeness with this “Phi” ratio matches with common list of most popular celebrities. Another example is the research published in the Hindustan Times newspaper on the smile of famous portrait of Mona Lisa2. As per this research on the movement of various face muscles of Mona Lisa in this portrait, Mona Lisa is having a happy face.
From the above two examples, we can see that if we do enough research in right direction, we can finally come to a conclusion which can be backed by numbers and hence science. Hence arts can be converted into science!
Since the start of the casting process, there is a mystery around complete process. Few of the casters / teams would get success and they would not reveal this success to others probably to avoid competition and hence keep it a trade secret. As the days pass by these secretes are passed on from one generation to another. Due to any reason, if this link / chain brakes, then success ratio in the process reduces and hence we say that it’s an art because only few persons can do it. In order to control and manage the complete casting process, we need to understand this whole process, break it into smaller and manageable segments so that we think scientifically about each segment to improve upon it. Let’s understand these steps with few of the examples below:
There was a famous company who used to make piston rings by casting a cylinder and then cutting slices of it. The average rejection rate of the finished product (to castings) of this company was about 50%, which was quite high. This company convinced to their clients about this is the average rejection rate in the process and cannot be reduced further at making correct sized piston rings is an “art”. They convinced their clients and hence their clients had agreed to accommodate the costs of these rejects into the cost of the Finished Goods. So there was no issues for the company about making losses (as they were financed by the client).
Once this company thought of going into quality certification, so the company appointed a consultant, who stumbled upon this unusually high rejection rate while making the piston rings. When he questioned about it to the company officials, he was told that piston ring casting process is very complex and its and “Art” and hence the rejection rates are very high. He wanted to go deep down the process of castings. So he studied:
After the initial study, the consultant realised that the average rate of rejection is almost same for all the days / shifts. Then he checked for the chemistry of the finished and rejected piston rings and found that the chemistry was also almost same and there was no abnormality in it. So even this reason was ruled out.
Diagram 1 – Principal of Centrifugal Casting Process. Courtesy “The Library Of Manufacturing”3
From his further study, the consultant found that the piston rings were rejected due to other defects such as inclusions, blow holes, etc. Then this consultant started investigating for those defects and wanted to check the process of castings. He went to the shop floor and started witnessing the casting process of various operators. Almost all of them were following the same procedure. So he could not come to any conclusion. He was about to leave when he saw one operator about to pour the casting. Before pouring, he was cleaning the spout of the ladle. This consultant was very surprised as he did not see this step of cleaning done by other operators. So he asked that operator about this extra step of spout cleaning. The operator told him that when he joined the company about 30 years back, his supervisor had told him to always clean the spout before pouring and he was following those instructions blindly (without knowing the repercussion of them). Later on when the consultant checked the rejection rate of castings produced by this operator, to the consultant’s amazement, the rejection rate was about 5% compared to the average rejection rate of 50%. So the consultant informed the company about his findings and the company started getting good results and off-course this added to their bottom line. Some of this benefits were also passed to their client organization.
Diagram 2 – Photo of cylindrical pipe making process (for representation purpose only).
From this example, we can conclude that totally innocuous action of spot cleaning rejected the rejection rate because every time we carry out casting, some slag or other impurities gets accumulated in the spout leading to their inclusion / blow holes in the casting and increasing the rejection rates. So if we find the individual rejection of the worker, we would say that this worker “knows” the art of pouring and other workers don’t have that art of pouring. But after systematic study, we say it was a science.
During the casting process, there are many parameters involved in each step right from storage of raw materials to the shipment of finished castings. If we break down this complete process into smaller manageable sub-processes and then further breaking down these sub-processes into activities, we could get some handle on it. Let’s see few examples in the art of casting:
While putting the above examples in front of you, I do not want to demean anybody’s experience or skills they have acquired while churning out heat after heat standing next to the furnaces in very hot / humid environment. The knowledge of shop-floor persons is wealth and my view point is how to acquire that knowledge and move from person dependent system to process dependent system.
Probably the above random examples would be very common in foundry industry and hence we always trust the skilled person for his skills and sometime yield to some of their demands (which may not be legitimate) as we have seen their results and their contribution to the company.
Let’s take one step back each of the above example and think rationally about them. This rational thinking should lead us to the desired solution. For few of us, who are experts, the following examples may appear to be very simple, however the idea is to think in the right direction to achieve the desired results.
Example 1: Incorrect calculation of Silicon / Manganese / other alloy addition. We all know that Silicon, Manganese, Aluminum, etc. have more affinity towards Oxygen than the Ferrous. Hence we use Silicon, Aluminum, etc. for ‘killing’ the steel or to “de-oxidise” it without generating FeO. So the XYZ melter uses this property of de-oxidizers to arrive at their correct addition without affecting the final chemistry. In order to do this correctly, he should know the dissolved Oxygen level in the hot metal bath and based on these levels, he can add various alloying materials containing the above elements in right proportion. While arriving at this conclusion, for simplicity I have considered that all other parameters such as bath temperature, chemistry and purity of the ferro-alloy, alloys, pure metals, etc. are same.
Diagram 3 – Cause & Effect Diagram for “Unpredictable Alloy Addition” leading to some causes
So in this example the XYZ melter gets very good results because he understands the importance of dissolved Oxygen and hence takes corrective actions to remove it from the bath. Others “see” the readings of dissolved oxygen but cannot correlated it with alloy addition. For additional details about this parameter, the readers can read research paper “Alloy Recovery and Control in Steel Melting”5 and “Optimisation of Ferro Alloy Usage in Steelmaking”5.
Example 2: High Alloy content in the next sample after alloy addition. Once the metal is poured into the ladle based on the chemistry of the sample, we add alloying elements and find that the alloy content in the ladle is much higher than the expected. If we carry out the root cause analysis of this case as follows then we can find the root cause
Diagram 4 – Cause & Effect Diagram for “Unpredictable Alloy Addition” leading to different causes
In order to mix these alloying element properly and homogenously in the hot metal bath, the temperature of the bath should be sufficiently above the liquidus temperature. Generally we know that the liquidus line between two alloying elements and accordingly we keep the temperature of the bath sufficiently above the liquidus line (not too high so that we can save on energy cost and not too low otherwise some bath metal starts solidifying). However the bath contains other alloying elements and the liquidus line no more depend only on 2 elements (which is the hypothetical case). Instead it has many more elements and this is especially true in case of high alloy steels wherein the percentage of each of the many alloying elements is significant compared to the base metal. In such cases it could be possible that the liquidus line of this multi-alloyed metal could be above or below the liquidus line of the two element system of iron with individual alloying element. If we do not have sufficient research data in this direction and it the alloying elements are not getting mixed up properly means that some part of the metal bath has solidified and we need to increase the bath temperature above the liquidus line so that all the alloying elements can get dissolved in the hot metal bath and we get homogenous alloy composition. For further details, the readers can read “Ternary Phase Diagrams”6. If we consider more alloys then the phase diagrams become very complex to plot and understand.
In this case it depends on 2 parameters one is the understanding of the melter about the situation and then increasing the bath temperature. For simplicity, I had considered all other parameters same.
Melting and alloying is a complex process and with the temperatures and current technology there are still many restrictions in fully understanding this phenomenon properly and accurately. So we have to depend upon periodic sample data and use our knowledge, skills and experience to arrive at the correct root cause analysis of the case to arrive at the right solution.
Solutions stated above have been over simplified so that all the professionals reading this article can understand the issues and methods to resolve them, so that we move towards making casting as a science.
+91-22-3598 3916 / +91-8355909335
