BHPV-The Lost Opportunities

There were many opportunities for BHPV to become a giant organisation but because the top management was timid and lacked imagination we let go of those opportunities. At least I know of three opportunities which would have catapulted BHPV to big league.
Let me narrate the first opportunity.
In early 1980's I was a member of the Polar Satellite Launch Vehicle (PSLV)design review team. Almost every month I used to go to ISRO at Trivandrum to participate in the design review. The review committee consisted of eminent people from design,materials,fabrication, testing etc. My job was to review welding and non-destructive testing procedures.
Rocket motor casing is nothing but a pressure vessel subjected to very high pressures. The entire design philosophy is to make the rocket shell as thin as possible but still meeting the safety and structural stability requirements. There were two teams who were designing the casing. It was understood that which ever team comes out with the best design in terms of least weight of the rocket meeting all structural stability requirements will be admitted. To give an idea of the intense competition between the teams, one team adopted the weld efficiency as 1.1 and another team adopted the weld efficiency as 1.08. At first glance this difference appears to be small. But this made a significant difference in the weight of the rocket motor casing. If one were to adopt weld efficiency of 1.08 one has to adopt a stringent welding procedure and an advanced method of detecting weld defects. The crack size was specified by the fracture considerations. My additional assignment was to find out the method of detecting the crack and if necessary specify an instrument measure the crack. The design review used to go from morning to evening and continued even during dinner time. it was the most stimulating experience i had.
I met several authorities in various fields and this experience was humbling.

During one of such meetings the Project Director of PSLV who knew me well called me aside and gave a bundle of drawings and told me that they were the preliminary drawings of PSLV and told me that he wanted BHPV to examine the drawings and give there consent for fabrication of Rocket Motor Casings. He also told that they want BHPV has to set up exclusive facilities for the manufacture and ISRO will completely fund the facilities. ISRO wanted air conditioned clean room facility, automatic TIG welding machines, assembling jigs, online testing facilities etc. which will also be funded by ISRO.ISRO wanted a guarantee that these facilities will not be used for any other jobs. I was excited for identifying BHPV for such a prestigious job. I told him that I will do my best to convince top management to accept the challenge.

What was the reaction of BHPV's management's reaction for this proposal will be the subject of next blog.

Establishing R&D in Bharat Heavy Plate and Vessels (BHPV)

In 1975 i joined BHPV as chief of R&D. Prior to my joining BHPV there was a small nucleus 3 post graduate engineers. Soon after I joined BHPV there was a change in the top management. I was reporting directly to the chairman and managing director and I had to wait for the appointment of chairman and managing director. After the new CMD took over he called me and asked me to go ahead and recruit M.Tech's from IIT's. We went to IIT, Madras,and IIT, Kharagpur. From these institutes we selected 8 M.Techs.4 M.Techs were from Chemical Engineering, 2 from manufacturing and 2 from welding. Till R&D building was constructed we worked from one single room. Our work shop was the the workshop of Training Centre which had primitive facilities. Within one and half years we had our own building. BHPV is a vast organisation with excellent fabrication facilities. But I soon realised that we cannot depend on the main manufacturing facilities for our prototype fabrication as the main manufacturing facility was not keen to divert their main machinery and workforce for our work. I argued with the CMD for our work shop. He granted us the finance to construct a work shop. The work shop and one shaping machine, one radial drilling machine,one milling machine, one lathe and several welding machines.Training centre sent us some trainee welders. We took them trained them again to meet our needs. When we needed shaping machine operators, milling machine operators, drillers and machinists the troublesome workers from the main shops were transferred to R&D work shop. Since we had no choice we accepted them. But within a short time we realised that these workers were really good. They were labelled"bad " because they always challenged established practices. We needed such people who can challenge existing practices.
In two years we had about 16 M.Techs in our group. 4 engineers formed the Chemical Engineering group, 4 formed the Cryogenic Engineering group, 4 formed the Manufacturing Technology Development group and 4 welding Technology Development group.
Each research engineer was allowed to work on 3 projects simultaneously and one project of his choice.Two engineers can join together and work on project when the project was big. When the project involved a prototype construction, the team may consist of one from chemical engineering who was responsible for process design, one was from product development to select material and method of manufacture and from welding technology group to welding methods and supervise welding. Thus a project team was a blend of various expertises.
Projects were selected based upon the organisation's need. We were constantly in touch with Sales to find out which ou the tenders the could not respond because of lack of technology. If such rejections are many, we would consider developing the missing technology. For example, at one time Sales could not respond to the customer inquiry for Titanium Heat exchangers. Acquiring the technology from abroad was expensive. Hence we initiated a programme to develop welding Titanium. We also went inside the workshop to find out the difficulties they are facing in manufacturing and initiated development projects to overcome this lacuna.
I also decided that at least a part of R&D expenses must be earned by accepting contract research. We approached, Defence Research & Develpment Orgniasation(DRDO)' Defence Metallurgical Research Laboratory(DMRL), Mishra Dhatu Nigam(MIDHANI)etc. for sponsoring projects. We were successful in obtaining a contract from DRDO to establish a manufacturing process for Titanium Gas Bottles for PRITHVY and TRISUL missiles. This project when completed successfully resulted in obtaining continuous orders of few crore Rupees. Similarly with DMRL for developing Armour welding and MIDHANI for Maraging steel welding.In these projects the entire development cost was funded by the respective organisations and also ensured that future business comes to BHPV. One of the contract clauses we insisted upon was that once the development was completed any hardware resulting from our development should be offered to us for series manufacture. If it was not aggreable, we were not we were not willing to take up contract research.
Another contract research we accepted was to design and develop compact heat exchangers for TEJAS supersonic aircraft being designed by Aeronautical Development Agency. The fabrication of the heat exchanger involved vacuum brazing of aluminium parts. In imported vacuum brazing furnace would have costed several lakhs of Rupees which neither ADA can afford nor BHPV can afford. We realised that a vacuum furnace is nothing but an externally pressurised vessel to contain vacuum and some heating elements and controllers to maintain predetermined temperature. We looked within BHPV and found that we were good at creating and maintaining vacuum, good at fabricating vacuum tight vessel and what we lacked was the expertise in control system for electric heaters. We decided that we will purchase the control system and integrate with the vacuum vessel and the heaters. We were successful in this gamble and we could build the vacuum furnace at a fraction of the cost of imported furnace. Not only this, we opened a new business of designing and building vacuum furnaces. Using this furnace we have brazed several aluminium heat exchangers. Later we built a high temperature vacuum brazing furnce for brazing stainless steel
When BHPV wanted to buy several kilometers of superinsulated pipe for transporting liquid oxygen and liqiid nitrogen from production plant to storage system BHPV found that the imported price of these pipes was busting its budget. We stepped in and offered to the Management that we could manufacture these pipes in R&D at a much lower cost. Mnagement was sceptical. We offered that we will build a trial pipe line selected by the Design department and subject the test pipe to any test desired by the QC department and if QC is satisfied with the performance, then we should get an opportunity build the entire pipe linr. R&D pipelines fully passed all the tests specified in the design document and we desined, manufactured and supplied the entire pipe line.
Another innovative idea I introduced was the transfer of technology to shops. I realised that a product or a process when transferred to the shop fails because it is not nurtured properly in the shops. Whenever a new product developed by R&D was transferred to the shops we insisted that the people who developed the product should also move to the shops. The engineer who developed it and the welders, fitters and others who worked in the projects were transferred along with the project. They can remain till the shop people absorbed the new technology successfully and felt confident to run it without R&D's help. We also offered that any of the staff so transferred desired to continue in the Shops can do so provided we are allowed to recruit new people in their place. This worked nicely for us.In one stroke we ensured that whatever we developed did not fail on the shop floor.
I think this is the first time any body thought of such a process of transfer of technology from R&D to shops.
A short list of various projects developed is given in my blog Dr. G.J.Guru Raja.

Starting and Running R&D

Before joining BHPV I was working in Central Mechanical Engineering Institute,(CMERI) Durgapur,West Bengal. CMERI was set up by Council of Scientific and Industrial Research,(CSIR). CMERI building was majestic and had very good infrastructure. It had overstaffed administration, finance department, purchase department etc. There was an elaborate system for purchasing any material. The system was so rigid that purchasing any material needed for Research was a tedious affair. Some times we had to wait for years before we can receive any material. Consequently Research workers had lost interest by the time the materials arrived. There was no system of review of the projects. Projects were started without identifying who is the end user. Projects were dropped without assigning any reason. Even serious minded people who were interested in Research got disillusioned and left CMERI. A quick analysis of the state of affairs revealed that the budget allocated was skewed. More than 70 percent of the budget went for maintaining the infrastructure. There was very little money left for actual Research. There was no accountability. Since the Institute was located in Bengal, non-Bengalis were discriminated. Studying all these at close quarters,I dreamed that at any time I had to lead a R&D group, I will allocate 70 percent of the budget to purchase machinery, materials, instruments etc. and remaining money for infrastructure. If necessary, have a separate budget for infrastructure and the budget for R&D will be decided by R&D staff. Also once a budget is allocated for a R&D project the spending cannot be questioned. There shall be continuous monitoring of the projects both for progress and budget.Any increase in the budget should be justified. Every project proposed should identify an end user and what benefit we are going to get once the project is completed. How some of these ideas were implemented in BHPV will be the subject of next blog.

Transport of Metal during welding Part-2

In my previous Blog which was published in May 2010, I had promised that I will write about the work done to explain under what conditions the molten weld metal moves backwards.

In 1970's A number of papers were published exploring how the crater is formed when an electron beam strikes a metal object. But we did not find any published work explaining how the molten metal moves backwards. As suggested by the Physics Professor in Moscow Power Institute we looked at the possibility of Thermocapillary phenomenon playing a part.

It is a well known fact that the surface tension force of metal depends upon upon the temperature. Higher the temperature lower will be the surface tension force. Therefore, in addition to the normal force acting on the molten metal a tangential force along the surface of the molten metal develops.The tangential force is determined by the surface tension gradient.This is represented by the equation
P(t) = grad(sigma)
A positive sign before the gradient indicates that the force P(t) moves the metal from a region of lower surface tension to a region of higher surface tension, i.e the molten metal will move from a region of higher temperature to a region of lower temperature which means from the front of the weld pool to the back of the weld pool.

When the electron beam moves over the metal surface the front portion of the crater is at a higher temperature and back of the crater will be at the melting temperature of the metal. As a result of the temperature gradient thermocapillary effect manifests it self. Hence under thermocapillary forces the molten metal flows from the front of the crater to the back of the crater.

The experiments reported in Part-1 confirmed that Thermocapillary effect is responsible for metal movement.

We also attempted to theoretically examine the metal flow.

To quantify the metal moved from the front to the back of the crater, neglecting the radius of curvature of the crater, as a first approximation, the flow of metal is treated as a two dimensional flow and relevant differential equations with boundary conditions were solved. The solution resulted in calculating velocity of the moving metal and the quantity of metal moved from the front of the crater to the back of the crater. The thickness of the liquid metal in the crater was calculated considering the temperature distribution in the crater. With this input we calculated the volume of metal moved from the front to the back. The theoretical calculations agreed closely with the measured values.

We were the first to identify thermocapillary effect as my Ph.D thesis was defended in 1972 and the work was done in 1971.

We published our work in Svarochnoe Proisvodstvo. Later my professor teamed up with the Professor of Physics mentioned earlier and together they published another paper extending the work to consider the flow as three dimensional. As our work was in Russian it reached not many people. But in spite of the disadvantage of language later I could identify 5 citations of the first paper published on Thermocapillary flow.

A practical application of this work was used to develop welding very thick aluminium blank of over 200mm in single pass by electron beam welding. Here the electron beam was located along the Y-axis as against the convention of beam impinging from top to bottom. In addition to the gravitational force, Thermocapillary force drove the metal away from the molten zone exposing virgin metal which aided in increasing the depth of penetration.

Transport of metal during welding-Part 1

Prof. Olshanskii suggested that I should examine what forces were responsible for driving the molten metal from the front of the pool to the back of the pool where the metal solidifies.
To visualise the flow of metal during welding I did some ingenious experiments. In a stainless steel plate of 10mm thick a hole of 2mm dia was drilled. Wires of copper,titanium,tantalum etc were embedded tightly in to these holes.A full penetration Electron beam weld passed over these metals embedded in the stainless steel.The welds were sectioned along the axis of the weld.The specimens were polished and etched to reveal the distribution of these tracer metals. The tracer metals had moved over a distance 6 to 8 mm in a direction opposite to the direction of welding. The concentration of the tracer metals along the depth of the weld and along the length of weld was measured using an electron micro probe analyzer.The maximum concentration was found at the top of the weld.

We were at a loss to explain how the metal could move such long distances considering that the dimensions of the electron beam welds are generally small.We searched entire literature to see whether any body had come across such a phenomenon. We could not find any thing similar to us.We realised that we have stumbled upon a new phenomenon. Finally my professor decided to consult one of the top class Physics Professor who was working in the same institute, that is, Moscow Power Engineering Institute. My professor took me to the Physics Professor and made me explain my experiments and also show him the results obtained till now. The Physics professor thought for some time and told me that I should look for "Thermocapillary Effect" and suggested that I should read a book on physics which was in the library. The meeting ended and we came out. I was disappointed with the meeting and I told my professor so. Then he told me something which I have not forgotten and which is helping me even now. He told me " I talked to several professors and all of them suggested that he is the right person to guide us. You may take a long time to find out the right person to consult. Once you have found him follow his advice implicitly. You jolly well go to library and read that book" I went to the library and read the book suggested by the Physics Professor. Indeed the back word movement of the metal could be explained by thermocapillary effect which will be the subject of next blog.

Dimensional Analysis applied to predict depth of penetration

The second portion of my thesis was to find an equation for predicting depth of penetration for a given welding parameter for a given material.

The formation of weld bead in EBW is controlled by many factors and to predict bead formation by rigorous mathematical analysis was not possible.While width of weld in EBW was not a significant parameter, depth of penetration is. Many research papers published at that time depended on fitting equations to the experimental data. These curves are applicable in limited field. To get a generalaised equation covering all metals and all welding parameter needed a different approach.

I had some experience in Dimensional Analysis and thought of applying this technique to predict depth of penetration.

The various factors that control depth of penetration are: beam current, beam voltage,beam diameter, welding speed and material properties such as thermal conductivity,melting point,specific heat,and latent heat of fusion. An advanced method of Dimensional Analysis was applied to solve the resulting equations. Three non-dimensional parameters were obtained. The first parameter related to the ratio of heat lost by conduction to the beam power, the second, the heat required to melt a given volume to the beam power and the last was the heat required to heat a given volume to the melting temperature to the latent heat of fusion. Plotting these parameters one against the other we got graphs which predicted the depth of penetration.The theoretical depth of penetration closely followed our Experimental results and the data obtained from other workers.These equations can used to predict depth of penetration in EBW for any metal at any given voltage,current and speed.

This was the first time any one had applied dimensional analysis to solve a welding problem.

Ph.D programme

After passing the Russian language exam which was conducted six months after joining the course, I was allowed to join the Department of Technology of Metals for my Ph.D programme. Taking in to account my background Prof.Olshanskii suggested that I study the temperature distribution during Electron Beam Welding and correlate the metallurgical changes that take place in the metal with measured temperature distribution. In addition he suggested that I should develop a method of predicting depth of penetration for a given welding parameter.

At first instance the problem appeared to be simple. But experimentally determining the temperature distribution had its own problems. Platinum-Platinum Rhodium thermocouples were used for measuring the temperature. The Heat affected Zone in Electron Beam welds are extremely narrow and to locate the thermocouples precisely in that zone was very tricky. Initial experiments were done to measure the width of the heat affected zone for various weld parameters. The welding chamber was equipped with a precision table which can move in the X and Y direction. The table was adjusted such that the electron beam passed through the centre of the table in the Y-direction. A jig was prepared such that the electron beam passes through the centre of the jig. The jig was bolted to the traverse table. Test specimens were machined such the they fit snugly in to the jig. Holes were drilled in the test plate and thermocouples were embedded to measure the temperature in the heat affected zone.The location of thermocouples was decided by the earlier experiments done to measure the width of the heat affected zone. The thermocouples were taken out of the welding chamber and connected to a high speed strip chart recorder. Welding was started and the temperatures were recorded. These experiments look simple but it was very difficult ensure that the electron beam traverses precisely where you want. Slightest changes in parameters or alignment of the test plate will result in the electron beam going over the thermocouples or going far from the thermocouples. The recorded temperature distribution was compared with the theoretical temperature distribution considering the heat transfer is two dimensional, The specimens were sectioned and were subjected to metallurgical tests.Several test specimens were welded to establish the repeatability. The first portion of the assignment was finished and the next was to find a method of predicting depth of penetration for a given welding parameter.