| KNURLING: The gap (or clearance) between two parts intended for brazing must be closely controlled to ensure that capillary action will draw the molten brazing filler metal (BFM) through the entire joint. This is not always easily accomplished, especially where tubular (cylindrical) assemblies are involved, because the allowable manufacturing tolerances on the inside diameters (I.D.) and outside diameters (O.D.) are often so broad (even if the parts have been turned or bored to fit), that they allow the fit-ups to be excessively tight or loose when one part is brazed inside another. To ensure that the inside member will stay centered within the outer member, manufacturers have resorted to many techniques for sizing, expanding, or contracting parts to provide for optimum brazing gap clearance at brazing temperature (the point where proper clearance is critical!). Simply pressing a very closely fitting member down into another (a "press-fit") might work if the diameter tolerances were exact, but this is often hard to accomplish; and for stainless assemblies, would probably cause galling of the parts. Knurling, however, is a technique that can help ensure successful press-fits. Knurling involves creating (by a variety of methods) long, vertical, side-by-side indentations around the O.D. of the inner member in the area to be brazed, so that when a press-fit is used for the two members, vertical capillary paths for the BFM can be ensured. See Figure 3.
Let's examine how knurling helped solve a brazing problem involving a stainless cryogenic valve body. Case History A valve manufacturer was modifying the design of a cryogenic valve for low temperature service at -313° F (-192° C) on liquid methane bulk-cargo tankers. Figure 4 shows a valve "bonnet," which consists of the top cover of the valve with a gland (a casting that surrounds an internal packing sleeve) which seals the valve stem. It is common practice to use an extension-tube in the design of the valve stem to protect the gland and actuator from extreme temperatures; however, this design had not been previously used on this type of valve pattern.
At the prototype stage, the decision was made to use a regular bonnet and gland casting, but to extend it by cutting the casting apart and inserting a stainless steel extension-tube to separate the bonnet from the rest of the casting (Figure 4). One reason was to avoid the lead time for obtaining new castings. Another was that the l.D.'s of the stem guide needed to be accessible for further machining before assembly. Vacuum brazing was the preferred method for joining the tube and casting. Conventional welding produced unacceptable distortion. A joining method was required which:
Vacuum brazing satisfied all these criteria and was able to provide for a necessary heat treat cycle at the same time. The joint was quite large: the prototype stem extension-tube had a 2.5-inch (63 mm) O.D. Thus, the circumference of the joint was almost 8in. (200 cm) long. A lab tensile-testing project had been conducted on a similar tubular joint brazed with Nicrobraz LM, the filler metal preferred by the valve manufacturer. Both gap control and centering were crucial to achieving maximum properties for this joint size. But, the machine shop was unable to maintain the tight tolerances necessary when boring the stainless sockets of the test parts. The unusual solution?-Relax tolerances! To accomplish this, knurling was tested.
A few small work pieces, which had used miniature knurling or splining to locate and center the plug-in-socket joints, had earlier been successfully brazed. Knurling was tested on a quick mock-up of this larger valve joint, and its sensitivity to varying gap sizes was examined. The results were surprising! A 1 mm pitch straight knurl was used on various diameter bars that had been intentionally machined slightly undersize. Knurling upset the metal locally, causing the extension-tube O.D. to increase (see Figure 3).However, provided the knurl was deep enough to allow a light pressfit, braze performance was very good over a wide range of fit-ups. Subsequent tests were conducted on the salvaged trial parts- which originally had tolerance problems when machined. The extension-tube had to be mounted on a gradually tapered mandrel to keep it round and maintain the diameter during knurling. The results were extremely good. Full-scale tensile testing of the brazed knurled assemblies resulted in breakage outside the joint, which proved that the joint was at least as strong as the base metal. Further testing showed that the range of diameter tolerances which could be allowed for successful brazing of knurled components could be increased six-fold over that previously used. See Table 1.
Brazing Procedure: Figure 5 shows the cross section of a filler metal loading groove, the volume of which was calculated to be 250%
of the brazing gap volume at the widest tolerance. Nicrobraz LM-S, a brazing
paste for syringe application, was liberally applied into the grooves and left to
dry overnight. The surplus BFM was scraped away, leaving a flush surface.
Components were then assembled with a fly press, and an internal mandrel was
used to ensure that they remained in alignment. The vacuum brazing cycle was
simple: a hold for uniformity at 1740° F (950° C), and brazing at1905° F (1050° C) for 30 minutes. Testing: Except for tensile testing on the trial parts, routine production called for cold-shock thermal cycling, halogen leak checking, and high-pressure bubble testing in water. No braze failures occurred. Conclusion: The new braze-manufacturing technique was so successful that more than 200 valve bonnets were produced using this design. See Figure 6. The procedures described have worked very well for this and many other applications; however, it is not a universal cure. You would need to test it for your own applications. This article is based on a case history assembled by Wall Colmonoy Ltd. (U.K.) for the High Temperature Brazing Support Group of the British Association for Brazing and Soldering (BABS). This information will soon be published as data sheet D-05 in a series of brazing case histories. For copies, write Wall Colmonoy Ltd. Peter Walter graduated from Cambridge University, after which he worked as a steel maker, and then became involved in the design and construction of large steel production and heat treatment furnaces. In 1970, he joined Wall Colmonoy Ltd. (U.K.), where he established the WCL stainless processing facility in Europe. Peter continued to develop the facility by installing small- and large-capacity vacuum brazing furnaces and providing support services. As Technical Manager, he initiated advanced ISO 9000 QA systems for both processing operations and materials production. Since 1990, Peter has been a consultant to the brazing industry. He continues to advise Wall Colmonoy Ltd. on QA systems, and represents their interests in surface engineering, materials and services. He is also convener for the BABS High Temperature Brazing Support Group and serves on European standards committees. Peter Walter's latest venture is the introduction of EBAS, Wall Colmonoy Ltd.'s European Brazing Advice Service. In cases where he can help solve brazing problems by phone or short-form fax, the service is free to all current or prospective customers of Nicrobraz products. More comprehensive services include on-site training, available per quotation. |
