Tuesday, June 30, 2009

superinsulsting cryo contaikners

In an ordinary glass thermos flask heat transfer from the hot or cold liquid takes place by conduction, convection and radiation. Heat is conducted by the neck of the thermos flak. Heat is lost by convection which is minimized by evacuating the system and radiation is controlled by giving a mirror finish to the outer wall of the inner vessel and inner wall of the outer vessel.In spite of these precautions heat is lost. These techniques are not effective if you have to store liquid nitrogen for prolonged periods. Liquid Nitrogen boils at -193 deg C.If the evaporation rate from the liquid Nitrogen container is 2 to 3 per cent per day conventional manufacturing techniques of glass thermos flask will not meet the evapolration rate as mentioned above. Hence an advanced manufatureing method is empolyed as described below.
A thin low thermal conductivity fiber glass tube is attached to to the inner aluminium shell by an epoxy resin capable of withstanding liquid nitrogen temperature. This fabrication method reduces conduction loss.
Multiple radiation shields reduce radiation loss. Multiple radiation shields consist of alternate layers of 5 micron polished aluminium strips 60 mm wide and low thermal conductivity cellulose paper wound on the inner shell in a pre determined fashion. The thickness of this blanket is determined by experimentation in a calorimeter to give minimum cold loss. A number of holes are punched through this blanket to allow air to escape during evacuation. The inner shell so prepared is connected to the outer shell by low thermal conductivity fiber glass tube.
Heat transfer through convection is reduced by evacuating the space between inner shell and the outer shell to a low vacuum of about 10 to the power minus 4 mm Hg.
A getter is used to take care of the out gassing that may occur during evacuation The entire assembly is heated from outside to aid degassing during evacuation.Evacuation may take one two days to reach the desired vacuum level. After checking the vacuum level heating is stopped and the container is allowed to come to room temperature. The container is checked for evaporation loss and the container is ready for dispatch to customer.

Saturday, June 20, 2009

Cryo containers

Operation Flood was conceived for increasing the availability of milk and milk products in the country.Artificial insemination of cows was a part of this programme.Semen from healthy bulls was collected, stored in liquid nitrogen containers and transported to various animal husbandry centers which in turn distributed it to various user centers.These portable liquid nitrogen containers are also called as cryo containers.
The cryo container is similar to a thermos flask.There is an inner container and an outer container. The inner container is connected to the outer container by a low thermal conductivity fiber tube. Just as in the thermos flask the gap between the inner and the outer container was evacuated to a low pressure. But there is one main difference.
In a conventional thermos flask space between the inner and outer container is evacuated to reduce heat transfer. The outer wall of the inner glass container and inner wall of the outer glass container is silvered to reduce the heat loss by radiation. But in spite of vacuum some heat is transferred by convection.Heat is also lost by conduction through the neck. In a cryo container all these losses are reduced to a low value by unique design which we shall see later.

Site stress relieving at Mathura Refinery

Our first opportunity for stress relieving of sphere was given by the management of Mathura Refinery. We had built a number of spheres for storage of LPG. Each sphere was 18 meters in diameter with a wall thickness of 65mm.
When we arrived at Mathura refinery our team was interviewed by the management to find out whether we knew our business or not. After satisfying themselves we were given permission to go ahead and start the work.
LPG was supplied by an LPG tanker. Since the withdrawal of LPG was very high natural evaporation from the tanker was inadequate.Hence we connected the LPG tanker to an evaporator which drew the liquid from the tank and the liquid passed through heat exchanger converting the liquid in to gas. Air for combustion was supplied by a blower. We had a stand by blower in case the blower failed during operation.
A number of thermocouples were fixed on the sphere and these thermocouples were connected to strip chart recorders. The entire sphere was insulated.We fired the burner out side the sphere to test its performance at various gas and air flows. After ascertaining the performance we fixed the burner at the bottom of the sphere. The top exit of the sphere was kept partially open for the combustion gases to escape. After checking all connections we fired the burner. The gas and air flow was adjusted to give a temperature raise of around 50 deg C per hour.As the temperature reached the specified stress relieving temperature the rate of heating was brought down considerably and once the temperature reached the stress relieving temperature the flame was adjusted to give a heat input which was just sufficient to compensate for heat loss from the sphere.After maintaining the temperature constant for specified time the heat input was slowly reduced and when the temperature reached around 300 deg C the burner was was switched off and the sphere was allowed to cool to room temperature. The cycle took nearly 20 hours.
Full credit of successful stress relieving should go to the head of chemical engineering and his dedicated team. Later a number spheres in Mathura Refinery were stress relieved using the same technology. We stopped requesting foreign help. More than 100 spheres have been stress relieved by now using the technology developed by R&D.

Wednesday, June 17, 2009

High heat flux and high velocity burner

Preliminary calculations indicated that the burner should meet the following conditions:
Heat output of the burner : I million BTU per hour.
The combustion must be complete.
The hot gases coming out of the burner must be given a tangential velocity so that the walls of the sphere gets scrubbed by the hot gases.
The flame must be stable at all conditions of air flow.
The burner must be easy to start in case of flame outage during stress relieving.
In case the air supply fails accidentally, the gas flow must stop immediately.

With such stringent conditions, we realised that we do not have the expertise in designing a stable high velocity high heat flux burner. We approached Indian Institute of Science, Bangalore and National Aerospace Laboratories, Bangalore for designing the burner. A development contract was placed with the understanding that the manufacturing of burner is our responsibility. Both the laboratories started the work simultaneously and with in one year we had two burners of different designs each meeting laid down specifications.

Meanwhile we developed the burner management system The burner management system consisted of an electronic system with infrared sensors to monitor the flame and activate a hooter in case of flame outage and shuts off LPG flow immediately. It also supplied high frequency power to the spark plug to ignite the flame. We did extensive trials with burner simulating all conditions of stress relieving and ensured that the burner can operate with out any problem. Then we were ready for site stress relieving.

Monday, June 15, 2009

How we developed technology for site stress of sphere.

In refineries and fertilizer plants liquefied petroleum gas and ethylene are stored at room temperature in spheres. These spheres are built from individual steel petals hot or cold pressed and assembled at site Depending upon the pressure and the diameter of the sphere the steel plate thickness can be as high as 65mm. The total weight of the sphere can be few hundred tons. BHPV,for some time had a monopoly in this business. After welding the sphere it has to be stress relieved. Stress relieving involves heating the sphere to a temperature of 580degC to 620degC depending on the plate thickness and composition and holding at this temperature for one two hours depending upon the thickness and cooling it slowly to room temperature. Since the sphere cannot be put in a furnace, the sphere itself is converted in to a furnace. A high velocity gas burner is fitted at the bottom of the sphere. The sphere is externally insulated and a number of thermocouples are fixed on the surface of the sphere which are used to monitor the sphere temperature.

The heart of the process is the high velocity gas burner. Initially a company from UK was contracted to carry out the stress relieving. This involved spending foreign exchange which was, before 1990 was not available easily and the logistics of getting the UK firm to come and perform stress relieving was very daunting.
The Management gave the assignment to R&D to develop the burner and stress relieving technology. We decided that chemical engineering group should handle this assignment.

Friday, June 12, 2009

conclusions

You have to remeber that all this development was done in mid 1980's.A research institute set up with UNDP help was publicising, at that time, that they have successfully repaired a tolling bell of a church and put it back in to operation!

Our success can be attributed to a team of young graduates who came to us with no baggage. They were willing to try any thing new. They were not afraid of failures as Management had realised that when you are working in high technolgy areas some failures have to be accepted.

The other factor responsible for our success was the working atmosphere in R&D. Whenever we met with a problem we used to call for a meeting of all people from various sections of R&D and the problem was thrown open for discussions. Welding engineers suggested solutions to chemical engineers and chemical engineers suggested solutions to manufacturing engeneers etc!. Some times we involved workers to suggest alternate methods of manufacturing. This approach helped us to think out of the box.

In the next blogs you can read the various pathbreaking developments we did in R&D.

Wednesday, June 10, 2009

Hydraulic testing of gas bottles

The acid test of the development of gas bottle is the hydraulic test,burst test and the stretch.
Out of the many bottles which had passed radiography, five bottles were picked at random by the customer and these five bottles formed a lot which are sent for hydraulic test. One bottle picked from this lot randomly is sent for burst test after hydraulic test.
The hydraulic test pressure is in excess of 350 atm. and burst pressure is in excess of 700 atm. Since these pressure are very high, for the sake of safety, the gas bottle prepared for hydraulic test was kept in a pit and covered with sand bags.The hydraulic pressure has to be raised in an incremental manner as given in the specification.We used a computer controlled hydraulic testing machine.The Incremental pressure raise and the final hydraulic pressure was programmed. Once the hydraulic test was started the machine took over and stopped when the desired hydraulic pressure was achieved. The hydraulic test continued to establish the burst pressure of the selected bottle. At the end of the test the hydraulic test pressure and burst pressure was read out from the machine. The full testing sequence was recorded and a print out could be taken at the end of the test.
Initially, when we used commercially pure titanium wire for welding larger bottle of 8.5mm thickness we had some failures.But once we switched over to 3Al-2V filler wire we did not have a single failure.
Hundreds of bottles have been supplied to DRDL for their use. Even today BHPV is the preferred vendor for these bottles.

Tuesday, June 9, 2009

Welding gas bottles

A trailing shield is essential for welding titanium or titanium alloy. In this method of welding certain amount of oxygen and nitrogen will be picked up from air by the weld.The pick up of these gases lowers the fracture toughness of the welds. For aerospace applications welding in open air is not allowed. Hence we used a chamber for welding.The welding chamber is a plexiglass dome mounted on a steel base.The chamber is connected to a vacuum pump which can evacuate the chamber at a fast rate.There are several feed throughs which allow current tobe carried to the welding torch, a welding positioner,an electric lamp etc. and a system of valves for gas inlet,gas outlet and evacuation of the chamber. The welder stands outside and performs welding through a glove port.
The chamber is first evacuated by a vacuum pump to a low pressure and then back filled with high purity argon.This argon filled chamber is again evacuated and high purity argon is filled for the second time.The chamber is maintained at a positive presuure by the flow of a small quantity of gas. The welder stands outside the chamber and manipulates the welding torch through a glove port and carries out welding
The machined hemispheres are sent to the metrology department to check uniformity of thickness and contour. If these are within the permissible limits,they are taken up for welding. Before the spheres are welded they are thoroughly degreased and the weld edges are etched with a solution of HF-HNO3. The welding wires are also degreased and etched with the above etching solutions.
Assembled hemispheres, filler wire and the other tools for welding are placed in the chamber before the start of the evacuation.To remove any residual oxygen and nitrogen in the chamber the welder welds on a scrap titanium plate. The welded titanium plate acts as a getter absorbing the residual gases, if any ,in the chamber. The welding operator standing out side lays the root bead by manipulating the welding torch.When the root bead is finished, the welded sphere is taken out and sent for radiography. If the radiography is cleared the sphere is taken for further welding in the chamber. For higher thickness spheres one more radiography test is performed at the mid thickness of the weld. These tests are done to detect any defects and rectify them and not wait till the sphere is completely welded. Specification allows rectification of the defects only two times. After the welding is completed the job is allowed to cool in the chamber. Then the job is taken out and sent for final radiography.If the bottle has passed the radiography test bottle is sent for hydraulic testing.

Saturday, June 6, 2009

Establishing welding procedure

After the hemisphere was machined,on each hemisphere one nozzle was welded in the specified location. These nozzles are specially machined to close tolerances.The nozzles have special threads so that once the gas is filled in to the bottle and closed by a plug the leakage from the gas bottle is almost zero.Two
hemispheres were tack welded and was ready for final welding.

We had to develop two types of gas bottles,namely,a smaller diameter gas bottle with a wall thickness of 5.5-0+0.5mm and a larger diameter gas bottle with a wall thickness of 8.5 -0+0.5mm

Customer had specified that the gas bottle should meet a certain stretch value at hydraulic test. This could only be achieved if weld test coupon had an elongation in excess of 8percent.
Welding procedure was established for a nominal thickness of 5.5mm and 8.5mm. When the test coupon was welded with matching composition filler wire.i.e,6Al-4V filler wire the desired elongation was not achievable.
When 5.5mm test coupon was welded with commercially pure titanium wire we could achieve the required elongation and the UTS as given in the specification. The weld bead picked up aluminium and vanadium from the parent plate which was responsible for meeting the required UTS and elongation.
8.5mm test coupon was welded with commercially pure titanium filler wire. The test coupon failed to develop the specified UTS. Detailed metallurgical analysis of the weld bead led to the conclusion that enough aluminium and vanadium was not picked up from the base plate as it happened in case of 5.5mm thickness plate. At the center of the weld bead the chemical composition was pure titanium and not an alloy of titanium.This was the cause for lower UTS observed. We realized that somehow we have to introduce aluminum and vanadium in to the weld without using 6Al-4V. We hit upon the idea of using two wires twisted together,one of commercially pure titanium and another of 6Al-4V and using the composite wire as filler wire for welding.When welded with this composite wire the test coupons passed all the tests.
We did extensive tests to ensure that the weld bead deposited with the combination wire did not lead to non uniformity of composition or gross segregation in weld bead.
We welded several gas bottles with this composite wire and the bottles passed all the tests.
Later the customer supplied TIG wire of composition Ti-3Al-2V which was used for welding large diameter bottles.

Friday, June 5, 2009

Machining the hemisphere

After the hemisphere was tested ultrasonically to ensure that there are no buried cracks, the hemisphere was sent for machining. Machining was done to removes the gas saturated layer and to maintain the desired wall thickness. Normally machining of hemisphere is done on a CNC lathe. We did not posses a CNC machine. But we had a brilliant engineer. He came forward and took the challenge of machining using the copying attachment to the ordinary lathe we had in the work shop.As the machining parameters such as speed,feed etc.are similar for machining stainless steel and Ti-6Al-4V he first experimented with a stainless steel hemisphere of the same diameter as the titaniun alloy hemisphere. In two or three trials he achieved the dimensions and shape. Then he took up the task of machining titanium sphere with the knowledge he had gained from copy turning stainless steel hemisphere.He could achieve the required shape as well as the required uniformity in thickness.We crossed one more hurdle. We were now ready for welding.
After the production process of machining was stabilized we invested in a custom built CNC machining center.

Thursday, June 4, 2009

Preparation of the blank

After deciding about the thickness of the blank, the next step was to prepare the blank for pressing.
The diameter of the hemisphere to be pressed decided the diameter of the flat plate required for forming. This diameter was further increased by about 10-15mm. The blank was cut by an oxyacetylene gas torch. When the blank is cut by oxyacetylene
gas torch two important points are to be kept in mind,namely, gas saturated layer and minute cracks that develop while cutting. Gas saturated layer increases the hardness and minute cracks can grow during pressing. Hence the blanks were machined till the gas saturated layer was removed and dye penetrant test was carried out.If no cracks are visible the blank is ready for pressing.
Elongation of Ti-6Al-4V increases with temperature.In the temperature range of 850-950 degC, the material will be in the super plastic range. Hence the blanks were heated in a gas fired furnace with excess oxygen to a temperature around 850degC.
When the blank is taken out from the furnace the blank loses temperature rapidly by radiation. When the blank is placed on the die for pressing the blank loses temperature by conduction. As the temperature falls the elongation of the material also falls. Our aim is to see that the blank does not fall below 500degC Hence to reduce conduction loss the top punch and the bottom die are heated to a temperature of 500degC approx.After ensuring that the dies have reached the desired temperature the heated blank was transferred rapidly to the press and the pressing was done without any loss of time. The pressed hemisphere was checked for approximate dimensions and was sent for shot blasting. After shot blasting dye penetrant test and ultrasonic test was carried out detect any cracks before the hemisphere went for machining.

Preparing the blank for pressing