Considerations for Orbital Welding in BioProcess Piping Applications

Editor's note: Pharmaceutical Online is pleased to present this four-part article on orbital welding for bioprocess piping by industry expert Barbara Henon of Arc Machines. This article was adapted from a talk given late last year by Dr. Henon at an ASME meeting.

ContentsWeld Schedules: Determination of Programmable VariablesPulse TimesCriteria for Weld Acceptance

Weld Schedules: Determination of Programmable Variables

RPM. Single-pass fusion tube welds are typically done at a travel speed of the arc at 4-7 in. per minute (IPM). The travel speed in in. per minute must be converted into rotational speed in revolutions per minute (RPM). Thus, for a 1 in. diameter (OD) tube, a travel speed of 5 in. would equate to 1.6 RPM.

Time per Level. The arc time of the weld would include the rotation delay time, plus the time to make 1 revolution at the specified RPM, plus the additional time required to travel a distance equivalent to twice the tube wall thickness to tie-in the weld. At 1.6 RPM the time required to travel once around the tube would be 60 seconds divided by 1.6 RPM or 37.5 seconds. The total time of about 40 seconds would be required to complete the weld. The total time is divided by the number of levels in the weld program to obtain the time per level. This would be 10 seconds per level for a 4 level weld or about 6.7 seconds per level for a 6 level weld.

Welding Current for 316L stainless steel. At a travel speed of 5 IPM it takes about one am-pere of welding current for each 0.001 in. of wall thickness for the starting (primary) am-perage in the first level. With pulsed arc welding, all of the other welding currents can be derived from the first level amperage. The background current would be about 30% of the pri-mary amps for level 1, while the current in the last level would be about 80% of the first level. This is the result of heat build-up in the tube so that about 20% less current is required for pen-etration at the end of the weld than at the start. The amount of reduction of current per level would depend upon the number of levels, with more gradual reduction possible with a greater number of levels.

Rotation Delay. After the arc is initiated, but before the start of rotation, the arc is held in one spot to build up enough heat for penetration. This is especially important for a single-pass weld where failure to achieve penetration at the start of the weld could result in a lack-of-fusion at the tie-in.

Pulse Times

The times for the primary pulse and background pulse control the spacing between the weld beads. Longer pulse times increase the bead spacing. In step or synchro welding, the "low" or "background" pulse determines the weld bead spacing, while the "high" or "primary" pulse time can be used in conjunction with the welding current to control penetration. On thin-walled tub-ing, welds can be made without pulsed current. For small tubes, pulse times are usually 0.1 - 0.2 seconds or less. Pulse times are significantly longer for STEP welds. For a pulsed arc weld, the weld beads should overlap by 60-80% on the OD and no less than 50% on the ID.

Criteria for Weld Acceptance

The ASME Bioprocess Equipment Standard (ASME BPE-97) was published in November 1997. Prior to this time, pharmaceutical pip-ing systems had traditionally been installed using the 3A Sanitary Standards as guidelines for fabrication. Welding procedures and per-sonnel may have been certified to the American Society of Mechanical Engineers (ASME) Sec-tion IX of the Boiler and Pressure Vessel Code and perhaps have followed the weld criteria guidelines listed in ASME B 31.3 Code for Pressure Piping by which the welds are visually evaluated to have no lack-of-fusion, no evidence of surface slag or porosity, with strict limits for incomplete penetration, undercut, ID concavity (suck-up or suckback) etc. To comply with these codes, a weld procedure must be established and test welds subjected to bend tests to show that the joint is ductile, and tensile tests to show that the weldment meets the minimal tensile strength for the material. Radiography may also be required. These tests were designed to de-termine the mechanical integrity of the welds and the capability of welding personnel to make the welds. These codes and standards were written for manual welding and it is perfectly pos-sible to install a piping system to meet these criteria using manual welding techniques. Welding to these codes is intended to assure the safe operation of the welded system, but little consider-ation is given to the cosmetic appearance and smoothness of the welds which will, in fact, affect the suitability of the piping system for biopharmaceutical use.

ASME BPE - 1997 was developed in recognition of the limitations of exist-ing codes and guidelines for the bioprocess industry. In 1989, the ASME es-tablished a Main Committee for Bioprocess Equipment to examine all aspects of bioprocess equipment fabrication and installation to define the in-dustry requirements and to write a new standard to address the particular needs of the bioprocess industry. Subcommittees for Design for Sterility and Cleanability, Surface Finish, Material Joining, Dimensions and Tolerances, Equipment Seals and General Requirements were formed and met several times each year until the Standard was completed. Work is still in progress on revisions and addenda.

The issue of weld acceptance criteria for orbital tube welds in bioprocess piping systems was addressed by the Subcommittee on Materials Joining. They agreed that biopharmaceutical welds must still meet the requirements for ASME Section IX and B31.3 but meet additional cri-teria based on visual evaluation. All welds are to be visually examined on the OD, and the number of welds to be inspected on the ID is to be agreed upon by the owner and contractor. A minimum of 20% of the welds shall be randomly selected for ID inspection, either by direct vis-ual examination or by borescope. It was agreed that pharmaceutical-type welds must be fully penetrated with no lack-of-fusion on the inner weld bead. In addition, there must be no evidence of porosity, nor slag nor dross, nor excessive discoloration due to lack of purge or too little purge gas, or contaminated gas, nor arc strikes. The MJ Subcommittee presented drawings of weld cross sections which define an acceptable weld profile and limits for concavity, discoloration, misalignment, etc. which are detailed below.

Incomplete penetration

Probably the most serious welding defect would be a lack-of-fusion or failure to achieve full penetration of the weld around the entire perimeter of the inside of the weld joint. Aside from considerations of the strength of the weld joint, in bioprocess piping applications a lack-of-fusion leaves a crevice where bacteria could escape the cleaning procedures and colonize the system. Crevices are also sites where crevice corrosion can begin. Differences in the microenvironment (oxygen, chlorides, metal ions, hydrogen) of the crevice and the area outside of the crevice set up a concentration cell with the crevice becoming anodic and hence corroded.

Incomplete penetration of the weld joint, or lack-of-fusion, generally results from a poor weld program where insufficient heat is applied during some part of the welding process. In this case, the condition can be corrected by increasing the amperage or welding current which provides the heat necessary to achieve penetration. The weld program may require additional amperage for all or for just a single part or level of the weld program. In general, when a weld schedule has been worked out for a particular size of tubing, or pipe, or tube-to-fitting weld, the welds will be consistent unless there is a change in material heat.

A lack-of-fusion may result from failure to align the tungsten electrode properly to the weld joint or arc deflection. This defect would not be apparent from the outside of the weld, but could only be detected by visual examination of the weld in-terior. This type of defect would be the result of "operator error" and proper training of welding personnel would be the most effective preventive measure.

Concavity (over-penetration)

With an autogenous weld, no fill-er material is added, so the weld surface will not be convex unless excessive purge pressure is ap-plied to the ID. The weld bead will usually be flush with the tube surface but on low sulfur or heavy-walled materials, there is more of a tendency to show some surface concavity which is considered to be undesirable. The outer weld bead usually becomes concave when excessive heat is applied to the weld. The con-cavity may be localized to one area of the weld, or the entire weld may be too hot.

Concavity can usually be overcome by reducing the welding current for the particular part of the weld joint showing the problem. On thick-walled or low-sulfur material, a slight amount of concavity may be unavoidable. The amount of permissible OD concavity for the new BPE standard is maximum 10% of the wall thickness for the entire weld circumference or 15% if the concavity is limited to 25% of the circumference. A 10% concavity on 0.065 in. wall pharmaceu-tical tubing would be measured as a 0.0065 in. depression of the weld bead with respect to the out-er tubing surfaces.

Excessive inner weld bead penetration

Excessive penetration of the inner weld (ID convexity) is limited to 10% of the nominal wall thickness. ID convexity usually occurs concurrently with OD concavity and is also the result of applying excessive heat to the weld. It can be corrected by reducing the amperage for that part of the weld showing the excessive penetration. While some owners or contractors specify a wide inner weld bead that will minimize the possibility of a weld having a lack-of-penetration defect, others want the inner weld bead to be as thin as possible and still be completely fused. If all of the welds in the system can be inspected by borescope, this may be an acceptable risk. Other-wise, it is much safer to accept a slightly wider bead and perhaps some OD concavity.

Discoloration or "heat tint"

The BPE-97 standard states that discoloration should be minimized on all product contact surfaces. The ID weld bead should be free of color, but a light straw color or faint blue color may be permitted in the HAZ. It leaves the fi-nal determination of the amount of color to be agreed upon by the owner and contractor. This is a controversial issue. The semiconductor in-dustry has long demanded color-free welds and the ISPE Baseline Guide also requires welds free of discoloration. Discoloration, or "heat tint" is undesirable because it is associated with a loss of corrosion resistance of stainless steel. A purge of inert argon gas on both the inside and the outside of the weld with good quality welding gas prior to, during, and after the weld is used to prevent oxidation. If no purge is applied to the inside of the tube, the welded area becomes black or "sugared." If a poor purge, or purge of insufficient time is used, discoloration ranging from deep blue to brown, tan, straw, pale blue or gray results.

The discoloration reduces corrosion resistance because the oxidation damages the passive outer surface of the steel, and oxidation of the chromium in the surface layer depletes the chromium around the grain boundaries which provide protection from corrosion. Perfectly pure argon gas (1 to 2 ppm O2 or less) and a perfectly clean weld joint should provide a weld with no visible evidence of oxidation. If the weld and heat-affected zone are free of oxidation, you can assume that the purging system was effective and the argon purity and flow rates were ideal.

Sometimes it may be difficult to eliminate all evidence of oxidation during welding. This calls for a fresh evaluation of the purging system. There should be no air leaks, and the tubing used to transport the gas from the cylinder or dewar should be completely impermeable to atmos-phere. Welded stainless steel is best, but Poly Flo (polyethylene plastic) is acceptable. Careful consideration of purge dams, end-caps, diffusers, etc., must be given. A highly purified source of argon and special filters or purifiers, such as the Nanochem or Gatekeeper which remove trace amounts of moisture, oxygen and other impurities from the welding gas may be effective in eliminating discoloration due to oxidation. The tubing itself may hold moisture on the inner surface which can cause discoloration during welding. Sometimes heating or baking the tubing can eliminate trace discoloration. A reliable oxygen analyzer that measures accurately in the low ppm range may be used to verify the purge conditions, but the ultimate test is color or lack of it, on or next to the inner weld bead when examined with a bright fluorescent light. The technology is available to achieve color-free welds on a routine basis. It is up to the end-user to determine whether the additional expense and effort to achieve a completely oxidation-free weld is justified for his particular application.

Failure to purge during tack-welding

ASME BPE - 97 simply states that all tack welds must be fully consumed. Tack welds are small spot welds usually done with the manual GTAW process but which may be done with an orbital welding machine. Tacking is done prior to welding in order to hold the pieces together for weld-ing. The safest method of doing a tack weld is to purge the ID of the weld joint in the same man-ner as is done for the complete weld. Oxidation on or near the weld joint could result in carbide precipitation, or otherwise initiate corrosion. In addition, the welding arc will travel in a straight line and consume a tack that is well purged, but may deflect around an unpurged tack which could result in a lack-of-fusion defect on the weld ID. It is also important to carefully clean the weld area prior to welding, and to use gloves when handling clean tubing, since oil or dirt on the hands is a carbon source that might contribute to carbide precipitation.

ID surface of stainless steel tubing showing appearance of purged and unpurged tacks. Center photo shows orbital weld deviating around unpurged tacks.

There may be reluctance on the part of installers to spend the money for gas, or another purge set-up for manual torch operations for purging tacks. Somehow they feel that if the (manual) welder is skilled, he can keep the tacks small enough so they will be easily consumed. Even when the tacks are small and don't penetrate all the way to the tube ID, some oxidation will still be trapped in the weld joint, so it's better to be safe and make the extra effort to provide a purge.

ID concavity

ID concavity is limited to 10% of the wall thickness by BPE - 97, but this must not re-duce the wall thickness below the design minimum thickness. With heavy-walled material, ID concavity may result from gravity acting on the molten weld puddle at the 6 o'clock position. For fusion welds of thin-walled tubing ID concavity may result from excessive pressurization of the inner weld bead by the purge gas which causes the molten metal of the weld to move outwards. It is possible to measure the pressure with a Magnehelic pressure gage. Purge gas from the tube ID is passed through the device and when the measured pressure exceeds 1/2 inch of water, there is a measurable displacement of the weld bead resulting in con-cavity of the inner bead. In extreme cases excessive pressure can result in a blowout where the liquid metal blows out and contacts the tungsten. This shorts out the arc, contaminates the tung-sten, and leaves a hole in the weld, and usually damages the weld head.

Concavity of the inner weld bead, also known as "suck back" is a defect. If it is severe, the effect can be similar to a lack-of-fusion defect, i.e., hard to clean, interferes with drainability, and of-fers a hiding place for bacteria to grow. The ability to displace the weld bead by pressurization has been used in some applications to provide a smoother surface than would otherwise occur. This is difficult to control accurately since the internal pressure changes during welding. Using this technique would certainly increase the cost of the installation.

Weld bead meander (doglegging)

Weld bead meander, or arc wander, is the failure of the weld to continue in a straight path around the weld joint, but to weave irregularly from side to side. If this irregularity is excessive, it can result in a lack-of-fusion which is a serious defect. The meandering can result from a damaged tungsten, from too high a flow rate of argon into the weld head which blows the arc around, or from dirt or grease or other contaminants on the metal surface of the weld joint, or contaminated gas. The arc may detour around unpurged weld tacks. It can usually be cured by a change of tungstens or by better cleaning of the weld joint. The most extreme examples of arc wander have been observed with the use of argon/hydrogen shield gas mixtures which often are contaminated and tend to be less effective in supporting an arc than pure argon.

Porosity

Porosity is voids or cavities formed by bubbles of gas evolved during welding. Porosity is more common in wire feed applications than in fusion welding. In fusion welding, it is typically caused by impurities in the base metal or from contaminants on the metal surface. Moisture ei-ther on the tubes or in the purge gas is a common cause of porosity. Porosity can usually be prevented by good cleaning procedures and by using purge gas of known purity.

Poor fit-up

Proper end preparation is critical to the success of orbital welding. The tube ends must be cut squarely and straight with no bevel. GF saws from George Fischer are frequently used as they can cut and square the ends in a single operation. Portable lathes such as those made by Tri Tool, Wachs, or Protem can provide a machined end preparation, but require the tubes to be cut first. Any burrs remaining after the end preparation must be removed without leaving a chamfer that would af-fect the wall thickness. The tubing ends must fit together without a gap in the weld head. The most common cause of a hole in the weld is poor fit-up.

Misalignment, or "high-low"

Misalignment of welded joints is limited to 15% of the wall thickness by BPE -1997. Misalign-ment of the tubes or other components being welded results in a ridge on the inside of the weld that can interfere with drainage of the piping system. Such liquid accumulation would promote the formation of rust and could lead to corrosion of the entire system. It can also result in failure to properly clean the system and permit bacterial growth leading to further contamination.

Misalignment can result from operator error in loading the weld head or from worn-out tube clamp inserts, or collets, that hold the components in the weld head. It can result from careless-ness in tacking, or from damage to the tubing or other components in shipping or handling. More commonly this high-low condition occurs when the fittings being welded to the tubes are made to different dimensional tolerances. If the tube or fitting is out-of-round, if the OD or wall thicknesses differ, this will produce some misalignment. You should be aware that the toler-ances for tubing are different from those for pipe. Misalignment is sometimes a problem when people try to weld pipe fittings, which have loose tolerances, to tubing which has tighter toler-ances. It is a matter of some controversy exactly how large a ridge it takes to cause liquid to accumulate and cause a problem. Both the dimensional tolerances of tubes and fittings and the amount of mismatch that can be present in the finished weld without compromising the drainablity of the piping system were actively considered by the BPE subcommittees. The Materials Joining Subcommittee has cooperated with end users to evaluate the amount of ridge on the tubing ID that would interfere with drainability and found that the presence of a ridge had less effect than expected.

Weld bead smoothness

Welds done for joining process piping for bioprocess applications must not only meet the crite-ria outlined above, but in order to achieve maximum cleanability must be exceptionally smooth on the inner weld bead. This is because bacteria are able to adhere better to a rough surface than to a smooth one. The criterion of a smooth inner weld surface is very difficult to achieve with any degree of repeatability with manual welding, but consistently good and very smooth welds are commonly achieved with orbital welding techniques. It should be mentioned that the smoothness of the weld bead is influenced to some extent by the quality of the material being welded. Valex Corp. has experimented with electron beam refined 316L material for high-end semiconductor applications. This steel is processed under very high vacuum and without the ad-dition of scrap material. No slag islands occur with this material and no discoloration at standard purge gas purities. There are no non-metallic inclusions and the weld bead is exceptionally smooth even at very high magnifications.

To see the previous installments of this article, follow these links:

I. Considerations For Orbital Welding In BioProcess Piping ApplicationsII. Considerations For Orbital Welding In BioProcess Piping Applications

For more information: Barbara Henon, manager, Technical Publications, Arc Machines Inc., 10280 Glenoaks Blvd., Pacoima, CA 91331. Tel: 818-896-9556. Fax: 818-890-3724.

Get the latest articles from Pharmaceutical Online delivered to your inbox.