Metal Casting Impregnation

Introduction

In the mid 1980’s casting impregnation came of age, when the US Navy invited our industrial members to be approved to their military specification – MIL-I-17563C. It was a landmark opportunity that the impregnation industry wasted no time in embracing. In 1997, due to financial restraints the US Navy announced that the specification would not be updated. This therefore places an onerous of responsibility on the impregnation industry, both suppliers of Sealant and Equipment, Applicators and OEMs to take the opportunity to build on the US Military foundation, a Quality Control Audit that embraces today’s needs. The requirement is for a comprehensive statement of intent in the impregnation industry to create its own quality control mantel that industry can undertake with confidence. This “Magna Carter”, thanks to the internet, can live in space, being available to anyone wishing to understand this very special technology.

A New International Standard for Impregnation

This institute has been created to foster excellence in impregnation technology. Its mandate is one of impartiality and transparency. The focus is to create an International Specification for Quality, into which all participating parties have a responsible part to play. The impregnation Industry has come a long way in the last 30 years and has established a professional standing in industry but it remains fractured with many individual standards. With this website, the Institute attempts to address some of these issues and place before you a set of rules by which a new standard can be established.

SPECIFICATION BRIEF

How to get the best from your impregnation plant

The main emphasis of quality control is built around the US MIL Test Ring. This is an excellent test to show how well the system is working. If the system is faulty, the ring can give some indication as to what could be the problem. As to the sealant itself, bear in mind that there are differing standards of sealants, some performing better than others. The US MIL Standard Test Ring is a useful tool when considering alternative impregnants as it is precise and will show even marginal differences between comparable products.

US MIL Test Ring

Initial questions:

  1. What type of sealant are you using?
  2. Is your sealant US MIL-I-17563C approved?
  3. Do you have an effective quality audit?
  4. Are you satisfied with the sealing performance?
  5. Do you use a test ring to evaluate process performance?
  6. Do you regularly impregnate more than once?

Checking Out The System

Sealant Condition

Ensure that your sealant is clean and free from contaminants such as swarf and similar debris. Remember that autoclave gel-ups invariably start with debris build up at the bottom of the autoclave/storage tank. So keep the system clean. Check clarity of sealant to ensure no water/oil is present.

Some sealants are more prone to contaminants than others. If there is discolouration or a translucent appearance to the sealant it would be prudent to request your sealant supplier test a sample for contamination. Some contamination of oil may not unduly affect sealing performance but could affect temperature and/or chemical operational performance. Trace where the contamination is coming from and eliminate it.

Check that sealant is being correctly degassed before use.

Degassing cycles are often overdone on the basis that the more you degas, the better the sealant. This is not the case. The sealant requires dissolved air as this is part of the stability mechanism. Removing this by extended periods of degassing will make the sealant anaerobic and cause it to cure. Absorbed air (as distinct from dissolved air) is the free air that is taken in from the atmosphere whenever the sealant is sitting at atmospheric pressure and this needs to be removed from the sealant before impregnation takes place. If not taken out, this free air causes significant exudation of sealant from the porosity during the curing cycle. The result is poor sealing.

The presence of this air can also have an inhibiting effect on the cure within the porosity. Link this with low catalyst level and you have a major curing problem. This free air generally comes off the sealant fairly quickly below 50 mbar and can normally be accommodated within the normal cycle time.

This applies to dry vacuum transfer systems only. Wet impregnation systems should have a separate degas cycle before impregnation commences.

Gel Time

Gel Time of the sealant is all important. Unless treating virgin components, the porosity is likely to be contaminated with a cocktail of cutting oils, water, emulsifiers etc., some of which contain anti oxidants. The latter can poison the sealant, especially if the sealant is low in catalyst thus making it impossible for the sealant to cure within the porosity. It also depends on the quality of sealant you have chosen as to the type and percentage of catalyst you can use, so please check with your supplier if you feel it necessary to exceed the recommended level of catalyst. Catalyst strengths in test tubes diagram.

All too often, catalyst strength is kept to a minimum because of the fear of the sealant curing in the autoclave. A cure in the test tube will only tell you that the sealant will cure in the test tube, but this is quite different to the cure in the porosity in both size and the risk of contamination in the latter.

Test tube cured sealant also benefits from exothermic reaction during curing. In comparison, it is very unlikely that any exothermic action would occur within the porosity due to the minute amount present. It follows therefore that non cure within the porosity is the greater risk. It is also important to note that once a sealant has been subjected to a cure cycle and remains uncured, it is unlikely to cure even if reheated. There is a greater problem here in that the uncured sealant within the porosity becomes a contaminant for further impregnations, rendering the casting scrap.

Therefore it is important always to have the maximum allowable catalyst present in the sealant. With little catalyst in the sealant, the slug in the test tube will appear stress free. On the other hand, sealant with the correct amount of catalyst may be found to be crazed and fractured. This should be ignored, as this condition does not relate to the sealant deposit within the porosity. It is interesting to note how the crazing in the test tube diminishes as the test tube gets smaller.

Sealant Storage

When working with higher levels of catalyst in the sealant, it is important to ensure that the temperature control system fitted to the sealant storage tank is working correctly to maintain permissible storage temperature to supplier’s recommendations. Bear in mind that a chilled sealant chills the casting resulting in a greater heat rise to the ultimate curing temperature. Also, to over cool the sealant will encourage condensation of moisture from the atmosphere which will be detrimental to sealant performance.

Check orientation of component – porosity position. See Graph Annex Two (rear of document)

As cast, casting will not seal. Incorrect position for impregnation Machined casting will not seal. Incorrect position for impregnation. This is often the greatest source of failure to seal. First, know your component. Know where it leaks. Orientate the component in the basket so there are no possible air locks around where the porosity is located. Do not make sealant draining the first priority; this can always be recovered in the drain tray. Failure to orientate the component correctly will generally cause the sealant not to penetrate and if it does, to be pushed out of the porosity, as a result of hydraulic impact, during washing and hot water curing.

Check that correct temperature is being applied for curing.

There are a good number of aging hot cure tanks in use today that have no means of circulating the hot water during the cure cycle. This can lead to cold spots within the tank and particularly within the batch of components. It is also possible that the heat input could be down for a variety of reasons such as caking of heating elements etc.

Temperature is a fundamental part of the impregnation process. No matter how good a sealant may be, if the heat is not there, it will not cure. Its mechanism is such that it’s either liquid or solid. The cure takes place in seconds but that short period is critical and the heat must be maintained to achieve polymerization. As indicated earlier, failure for the sealant to cure when first heated, it is then unlikely to cure even at a higher temperature.

First ensure that the heating is adequate for the load of components going through the plant. If not, reduce the load accordingly. Maintaining a large load and extending cure times does not work. Run the water cure at maximum temperature and ensure that the heat loss during the cure cycle does not drop below 90ºC. In terms of heat circulation, fit a powerful centrifugal pump with filter or other device. Do not fit aeration as this removes heat.

Your tank should be regularly checked with a portable digital immersion probe in a number of areas, especially close to the bottom, and compared with the resident sensor. It is important to know what processes, if any, the component is going to be subjected to ahead of the point of carrying out the impregnation process. As mentioned earlier the cocktail of chemicals trapped within the porosity can have a dramatic effect on whether the sealant cures or not.

Stress relieving from the foundry can be important. Some heat treatments call for oil quench, some call for hot water. Needless to say the latter is preferred. Avoid oil at all cost – porosity saturated with oil is a non-starter. Once the oil has penetrated the porosity, it cannot be removed by vacuum alone. This means that if oil testing is required, it should be carried out after impregnation.

Preparatory Drying

Oven drying has always been the traditional method of preparing the component by, cooking the components at around 150ºC for one to two hours. Often this procedure is the first thing to be dropped when operations or operators are under time constraints. Again, there is no way of telling whether a component needs drying or not. Making light of decision making – if it doesn’t look as if it needs it why do it?

For some time we have been aware of the virtues of vacuum drying and its benefits over oven baking. The presence of both temperature and vacuum allows significantly greater scope in the removal of highly volatile matter such as that used in cutting fluids. The dynamic effect of differential pressure across the cross section within which the defect is sited also has the effect of forcing out liquids from the cavity that would not otherwise budge by the oven method.

As an example: To vacuum dry a batch of components preheated to say 80ºC and vacuumed down to 20mb equates to 340ºC in an oven at atmospheric pressure. This is not to say that desirable results cannot be obtained at lower vacuumed temperatures. With aluminum and plastics playing an even greater part in our technical advancements such an oven temperature as this would cause havoc and be unacceptable to component requirements. Needless to say that there are instances where the volume of water within a cavity exceeds the latent heat stored in the surrounding metal or composite material would have the effect of freezing of the entrapped moisture. In such cases, oven drying would be more appropriate. This should not necessarily be a problem with powdered parts as liquids can and should be avoided with effective production planning.

Vacuum for impregnation

As distinct from vacuum drying, the vacuum cycle (dry vacuum) for impregnation can be short especially when carried out at elevated temperatures. This is specially so for virgin components that have been kept clean and dry. Generally a vacuum of between 20 and 10 mbar is adequate for most applications. This ensures that the free air has been removed. Anything below 10 mbar would appear to have little effect other than removing absorbed air and monomer from the sealant.

Using test rings in routine processing

If your sealant supplier is approved to US MIL-I-17563C, the most effective tool for quality audit is to pass a US MIL-I-17563C Test Ring through your plant. Check with your supplier for availability. Obtain certified rings if you can.

Identifying Correct Test Rings

Due to its composition the US Mil Test Ring is very difficult to manufacture and therefore expensive to use for routine testing. A substitute ring having identical porosity levels has been in use for many years that has proved most useful as a Quality Assurance Tool. Ask your supplier if he can supply these rings or perhaps advise where you can obtain them from. Please contact us if you have any problems.

Using the US MIL Standard Test Ring to test overall plant performance

Placing the test ring or several test rings in different parts of the process basket and passing it/them through the plant may also reveal some interesting results. It allows you to know more about your plant and where possible problems lurk – forewarned is forearmed. It is the one certain standard against which performance can be judged. It may not be able to tell what is wrong with the process application but at least you will have the knowledge whether the process is working correctly or note.

Testing with US MIL-I-17563C Test Rings

  1. If possible, first weigh the test ring to 0.1g precision and record.
  2. Secure the ring to a component about to go through your plant.
  3. Carry out a normal impregnation cycle, washing and curing.
  4. Remove ring and re-weigh and record. Well applied high quality impregnant will increase the weight of the ring by as much as 1 gram.
  5. Clamp into pressure test fixture and apply air pressure of approx 3.5 bar. DO NOT SUBMERGE IN WATER AT THIS STAGE.
  6. Observe the ring when pressure is applied and note if anything exudes – this is likely to be uncured sealant.
  7. If there is no such exudation, immerse in water and observe any leakage.
  8. If the part should leak:
    1. Compare rate of leakage with the leakage chart shown Appendix One (rear of document) and record.
    2. Return the leaking ring to the plant and carry out one further impregnation, test and observe again as above.

In accordance with the US MIL-I-17563C approval, the ring should be sealed completely. The US MIL spec allows for a second impregnation to seal. In reality, a second impregnation should only be necessary where the component is of a poor grade or the system is failing. If the sealant is of good quality and the process is correct, you should get a seal from one impregnation.

Note: Even high quality sealants will fail if the method of processing does not meet the laid down specification.

In Conclusion

The component might seal but then how reliable is it? This is down to the application and chemicals used. How many in industry, who have impregnation carried out, understand the process sufficiently to ensure that they are getting what they expect? The answer has to be ‘almost none’. There are no international standards to speak of, so this leaves it open to individual companies to produce their own – of course favouring the way they do things which can be meaningless to the end user. If a component fails after the process who knows whether it’s the treatment or the casting that’s at fault? Very often the innocent and indefensible casting gets the blame!

The answer can only lie in the integrity of the process application as a whole.

The golden rules:

MIL Standard Review

United States Military Specification Impregnation MIL-I-17563C

The US Military Standard initially appeared in 1985. Whilst PDY first employed sintered test rings in 1965 for both sealant development and process control, this was the first international standard of its kind employing a sintered ring for the purposes of impregnation sealant examination and approval. From the impregnation industry point of view it represented a major step forward by giving some credence to this little-known process with a standard that was considered worthy of international cover. A number of companies supplying sealants to the impregnation industry have over the years signed up to this standard which is now increasingly used as a symbol of quality by suppliers and users alike.

Predominantly, the standard is written to include approval of both the impregnant and its application. The standard is well written and informative as well as it can be, having to cover an application that cannot be seen and measured. It follows therefore that there remains a void between approving the actual impregnant and the actual application of the process in the field. On the one hand the testing of the impregnant by the sealing of a test ring is simple to define, the ring either seals or it leaks. This is easily ascertained after impregnation by putting the ring on pressure test and observing the pressure tightness of the ring. However what it does not tell us is, why the ring has sealed or not sealed. We do not have any knowledge of the condition of the porosity, it could be that it would help or impede the sealing process. It does not tell us if the ring was processed correctly. We just cannot see what has happened.

In comparison, we can take a piece of metal and turn it down to a particular size and measure to ensure that this has been achieved. We can analyse the metal and ensure that it meets the intended specification. But with impregnation we can only assume. So what does the US MIL Standard do for us? Well it tells us that the sealant is fit for purpose and will effectively seal components that fall within the scope of the test ring and that objective tests can be carried out with these impregnated rings to ascertain immersion chemical and temperature resistance. So far so good! What the present approval compliance is not able to tell us is:

  1. That test rings were of a set standard in density and uniformity.
  2. Whether the sealant submitted for the evaluation is identical to that which is made available to the market place.
  3. Whether the sealant has any identity verification or traceability.
  4. Whether the test rings were limited to two impregnation cycles.
  5. How well the sealant performed in quality and performance.

Test Rings

The standard states that it is reliant on the applicant sourcing and providing their own rings for the test approval. Invariably, these rings are made in small quantities, by different manufacturers and at different times which must lead to questionable standard of uniformity. It has been experienced that some rings were found to be so inconsistent and poorly made that they were potentially dangerous when it came to pressure testing. No doubt at the other end of the spectrum there are rings that were found to seal with ease i.e minimal porosity. Such rings, therefore provide an unstable and unsuitable platform for reliable evaluation.

Sealant

The applicant is conveniently allowed to seal the test ring himself. However, there is no suggested method to allow verification to show that this has in fact occurred. An applicant could easily seal the test rings with another supplier’s sealant, known to comply with the specification without fear of detection. Furthermore, there is no means of confirming that the sealant being used in the field is one and the same as that approved to the standard.

Impregnation Limit

As the applicant is relied upon, at his own discretion to impregnate to a limit of two impregnations, again there is no way of verifying that this has been adhered to. Bearing in mind, having to deal with questionable standard of test rings that in themselves may not be taken for granted, the test assessment begins to look somewhat insecure. It must be assumed therefore that some licence in the sealing of these test rings would be conceivable although not in the spirit of the requirements of the specification.

With the absence of traceability and particularly having to cover a process that cannot be seen or measured, its application is clearly open to abuse, undermining the very foundation of credibility of the MIL-I-17563C Standard. The gulf between that which is approved and that which is being applied in the field, even if the original approval was legitimate, does not relate with the specific needs of the specification. For instance concern over reactive substances (MIL-I 3.3 Volatility) can be brought into question because we are now talking about a sealant that is in use and contaminated. This contamination can come from incoming components for processing of which origin is invariably not known. Therefore any contamination that it may harbour may be left in the sealant to contaminate subsequent components.

There can also be a vast difference between the scope of controls over the application of the process employed by the user as opposed to that exercised at the time of the preparation of the test rings for MIL approval. Many factors can influence the performance of the sealant. As has already been highlighted, the level of catalyst in the sealant can have a major influence on how the sealant behaves in the porosity. Because high levels of catalyst can breed fear of pot life instability, there is a tendency for process plants to operate at the lower end of reactivity which can lead to the sealant not curing or only partially curing within the porosity. If we add possible contamination such as water, oil, wash solutions etc., the sealant becomes even further removed from the laboratory prepared sample supplied for MIL approval.

It should be further understood that any liquid, vacuum impregnated into a porous component can exhibit a level of sealability to that component. Even water can provide a sufficient seal to get an otherwise leaky component through production pressure test. Providing of course that the duration of test is shorter that the time required to purge the water from the porosity.

As we can see it is becoming increasingly difficult to relate to that which was approved and that which is actually in use in the field, making the reality of the test procedure adrift from that in the real world. So what can be done about it?

Quality Performance Of The Sealant

Unbeknown to the market place, there can be considerable difference in quality and performance of impregnants. The present specification groups all approved sealants into one, fusing both high quality and poor quality together. Certain tests can be introduced so as to filter out the various layers of quality. This information is equally important to the end user of impregnation as it is to the intended applicator. The question of explosives is a point in question. Should a low quality sealant, able to pass the present approval tests be chosen for ordinance processing. First we need to understand the difference between a high quality sealant and a poor one.

A high quality sealant must:

  1. Exhibit good long term Pot life Stability.
  2. Be stable with fast gel time – right amount of catalyst.
  3. Be able to cure correctly in the presence of reasonable levels of contamination.
  4. Have low viscosity and low surface tension for good penetration.
  5. Be easily removed from the component surfaces prior to curing avoiding the need for high shear washing action.
  6. Maintain test ring pressure tightness after elevated dry air temperature exposure.
  7. Have good all round chemical resistance.

Recommended Addition/Change To US MIL-I-Standard

  1. Test Ring – (Aluminium)

    Establish a central supplier of the test ring. Consider a test ring that is manufactured from commercially available powder. A ring that has a high degree of uniformity and repeatability.

  2. Gas Chromatography Verification

    For the applicant to provide a representative sample of the sealant to be tested by GC/FD. This provides a footprint of the product that can be held on record on an approved website. The applicant’s confidentiality is not breached as such information could easily be available in the public domain.

  3. Test Ring Preparation

    For the central laboratory to acquire its own test rings from the approved central source and carry out the impregnation with the applicant’s sealant according to applicants publicly declared information. Test rings to meet the existing requirements of MIL standard – max 2 impregnations. If the rings fail the sealant is rejected.

  4. Sealant Peroxide Test

    Before commencement of the MIL Standard application for approval, sealant containing AZDN as the catalyst should be checked for the addition of peroxide catalyst. This is known to enhance poor sealant performance. However such mixing of catalysts is also known to cause pot life instability if used in a production environment. A simple test is available for detecting the presence of peroxide.

  5. Gas Chromatography Chemical Testing

    MIL 4.5.5.6 requires that sealed and pressure tight test rings are subject to Table 111 and shall remain pressure tight when tested to 50 psi. Such a test does not examine the possibility that there could be some migration of the sealant into the test media. This could take place even if the test ring remains sealed for the duration of the test. To be certain that such a condition does not exist, GC/FD examination of the test media for elements of the sealant should be carried out. To overlook such a potential situation could lead to problems in the field.

  6. Sealant Contamination

    Sealants can be in use for many years and expect to become contaminated. Well designed sealants are able to cope with reasonable levels of contamination and still perform satisfactorily. It is therefore recommended that the MIL Standard test procedure incorporates a contamination test containing a mixture of a regular aqueous wash solution at 5% by volume and mineral oil at 2%. Antioxidant in these products can have a significant effect on curability of sealants that have low catalyst content.

  7. Raw Material Contamination

    Sealants based on methacrylate monomers can sometimes contain traces of polymer as a result of an over reaction manufacturing process. All incoming raw materials for sealant formulation need to first be checked for polymer contamination. The presence of polymer can effect proper penetration of the porosity. This is a test that can be carried out in conclusion of – MIL Ref: 4.3.1.1 Pot Life Test.

  8. Quality Grading Of Sealants

    It should be possible to award points to the testing of a sealant. These points would be collated to provide a star rating **** for high quality and * for low quality. These gradings could be on view on the information web page for impregnants. The design authority would have access to this information and thus be able to make a distilled choice based on particular needs.

    As an example:

    1. Record whether a sealant requires one or two impregnations to seal the test ring.
    2. Record final sealant weight pick-up and sealability of the test ring.
    3. Record weight loss of sealant after 24 hrs at 150ºC.
    4. Register rates of ring leakage, using the ring chart as noted throughout the test evaluation.
    5. Grade test performance of contaminated sealants.

Again the availability of such information to the market place will encourage effective quality assurance at all levels. It will allow the original MIL test approval to relate to the actual application of the process.

Test Laboratory Responsibility (Recommended)

Conclusion

If an impregnation standard is to have any meaning it must relate to the actual application of the process in industry. The key to this must be the test ring. It is the one thing that draws the whole process together.

It is appreciated that there may be good reason why the existing US MIL Standard Ring is required to be manufactured from pure aluminium. However, from a commercial point of view it is more important for the ring to have regular conformity in its density and available porosity.

The use of a GC/FD trace allows traceability between the original MIL Approval and the product supplied to the process plant. It also allows the user of the sealant to ensure that the sealant is not compromised. If there is any doubt, the user should have access to the original sealant trace against which he can check the sealant’s identity. Such a check could be carried out at the appointed test laboratory. A regular trace of the sealant might also provide an early indication of contamination.

Such available information can also be useful in the event of a dispute between supplier and user of sealant. It also provides a useful base on which to apply an international standard, one that can be adopted by most impregnation users for internal quality audit. Often sealants are purchased primarily on price, because it is the only visible means by which it can be commercially judged. At this point in time the US MIL-I-17563C is the only means of judging sealant quality, which if it remains ineffective in its present form could provide a poor image of the impregnation industry. We have a collective responsibility to approach the US Military and make our recommendations known and request that this standard be updated to a point that it becomes an accredited document with audit traceability that both supplier and user can refer to with confidence.

It is important to bear in mind that once the sealant has been decanted from its drum into the process chamber it is in a totally different world from that of a laboratory environment. The mere presence of metal has a considerable influence as to how the sealant behaves. Unlike paint that is mixed and used once, sealant may be in use indefinitely.

Frequently Asked Questions

When should I impregnate?
This depends on where the leakage is; from a cast surface, or from/to a machined face. If the former, it is advisable to impregnate in the as cast condition. Otherwise, it’s better to impregnate after machining.

What size of porosity can be sealed?
This is a question everyone asks. Unless you can see through the defect, it is impossible to judge. It’s not what you see on the surface that matters, but the formation of the porous channel through the body of the component. Modern day methacrylate impregnants, due to their low viscosity and the very nature of the impregnation application, used in the process is being unable to seal seriously large defects and especially if structural weakness exists. The best advice is: If in doubt, impregnate it. It is best to allow the process to decide what can and cannot be impregnated.

What types of materials can be sealed with methacrylate sealants?
Thermal cure sealants are not normally affected by different materials such as iron, bronze, aluminium, ceramic, GRP,
carbon. In fact any inert material that is porous is impregnatable.

How can I guard against sealant failure?
Know your component – identify where it leaks. Place it in the process basket so that it does not air lock. Regularly check your sealant and most important of all check your cure temperature. The latter is one of the most regular reasons for component failure.

How many times can I impregnate?
Assuming your system is working correctly, one impregnation cycle only should be all that is needed. Twice should be the maximum. With a correctly run system and good quality sealant, it is not usually necessary to impregnate twice. Check your system and quality of component if the latter is common place.

Should I impregnate components cold or warm?
It has been found that it is beneficial to impregnate components warm, but check with your sealant supplier first as some methacrylate sealants are more sensitive than others to increased ambient temperatures.

How long can a methacrylate sealant be used for?
High quality sealants can be used almost indefinitely, dependent on conditions of application. Good housekeeping is
important, particularly ensuring that components are clean and dry and free of contamination.

What is the difference between wet and dry vacuum impregnation?
Wet vacuum impregnation is the workload is directly placed into the sealant within the autoclave and a vacuum is pulled through the sealant. When using dry vacuum impregnation the workload is placed into an empty autoclave, the vacuum is achieved first then the sealant is admitted to the autoclave before returning the autoclave to atmospheric pressure.

What are the benefits of sealant recycling?
There are many, but most important is eliminating 85% total loss of sealant down the drain and its impact on the
environment. Recycle sealants have been in use now for almost 20 years, having achieved full US MIL 17563C
approval. Sealing performance is ranked high in comparisons with traditional sealants. In addition, constant recycling
ensures the sealant is maintained in a clean usable condition.

APPENDIX ONE

Test Ring Chart

US Mil Standard Test Ring Mil-I-17563B refers.
Porosity level 18-22%

Note: Some sealants may be found to not be able to seal at the sealing faces with the “O” rings. Such leaks, if bad enough to mask the inspection of the body, can be minimised by first rubbing the ring ends with wet/dry fine paper on a flat surface. Seal face leaks should not be counted in the scoring. If the ring is found to be completely sealed, inclusive of sealing faces (without rubbing the ring ends), to be scored as “00”

Rings & Test Fixture supplied by X-Seal Ltd., PO BOX 357, Lymington SO41 8WF

APPENDIX TWO

Fig 1 As cast
Casting will not seal
Incorrect position for impregnation
Fig 2 Machined
Casting may not seal
Incorrect position for impregnation
Fig 3
Correct position for impregnation
Fig 4
Correct position for impregnation