Click here to upgrade to a newer version of Internet Explorer or Microsoft Edge.
The Welding's Done! Now What?
By Bob Whittier (originally published in EAA Experimenter, July 1999)
You’re one of the participants in a somewhat dull hangar flying session,and suddenly the urge strikes you to liven things up. A great way to do this would be to say, "I’ve finished welding my homebuilt’s fuselage and am wondering whether or not to normalize it?"
Some of the members present will look at you quizzically, having no ideaat all what you’re talking about. And the others will all start yakking excitedly, spouting as many different opinions as you’d hear in a debate between political liberals and conservatives.
The subject of what to do to steel after welding it is, in all probability, more misunderstood than any other aspect of amateur airplane construction. The problem stems from the fact that over the years, so many people have used the terms "heat treating," "stress relieving," "normalizing" and "annealing" so loosely and incorrectly that they’ve collectively but unintentionally built a technological Tower of Babel.
While researching the articles on electric arc welding methods that appeared in the May and June issues of this magazine, we soon noticed that various books on welding had markedly different things to say about these subjects.
Some books written for student aircraft mechanics say nothing at all about normalizing. There’s no mention of it in the FAA’s guidebook for aircraft mechanics, AC (Advisory Circular) 43.13-1A , nor in the new revision 1B. A Navy manual on aircraft maintenance published in 1950 does not use the word "normalizing" but instead mentions a process it calls "drawing off," and then fails to explain what this means. An Air Force manual dated 1963 says that normalizing can only be done in a heat treating furnace in which both the temperature and atmosphere can be precisely controlled. That’s certainly not much help to a mechanic ordered to perform a welding job outdoors and in midwinter.
Bear in mind that some textbooks and manuals are compiled by publishing house staff members who are not engineers or mechanics. Some are written by experts and proofread for technical accuracy by other experts — but not by newcomers for understandability. Some technical books are written by people who are very impartial and objective; others by people who fill their copy with their own and sometimes contentious opinions. In the course of our reading we encountered a fair number of erroneous statements and a few that were preposterous. And so on.
It’s little wonder that today’s homebuilt airplane enthusiasts are confused. Their confusion is compounded, sometimes to the point of alarm, when upon talking to assorted kit plane manufacturers they learn that some of them " normalize" and some don’t. We came to feel that if the authors of several different books on welding were to be locked in a room to settle their differences about the matter, a fist fight would soon erupt.
So what is poor, confused Joe Homebuilder to do? Well, we’ll try to be informative in this article, but we can’t promise to offer the kind of quick, pat answer that many people today seem to want.
A piece of steel held in your hand certainly looks and feels like a very solid and homogeneous material. But when a suitably prepared specimen of it is put under a powerful microscope, it is revealed to have a crystalline structure. Library encyclopedias, manuals on steel and manufacturing processes and texts on metallurgy are full of microphotographs showing the vast number of ways in which this structure varies depending on how different kinds of steel are alloyed and heated. Sometimes artist’s drawings are used for this purpose, for such reasons as showing certain things to best advantage.
When shipments of aircraft steel tubing arrive at our homes, we quickly rip open the containers. We take the steel itself for granted, and happily set to work sorting, cutting and welding it. Homebuilt planes have been built and safely flown with no more knowledge of this metal than the fact that it’s of aircraft grade and that the builder has a passable level of skill with a welding torch.
Nevertheless, it helps greatly in using this metal to best advantage to understand and appreciate it more thoroughly. We can start by noting that metallurgy covers a vast field and one, therefore, can’t learn it in a few evenings of reading. Humans have been working with iron and its derivative steel for at least 5,000 years, and collectively have learned very much indeed. Some universities offer majors in the subject. Some of the metallurgists employed by steel and welding equipment manufacturers hold Ph.D. degrees. Like aviation, this field of endeavor makes use of a great many specialized technical terms. The best we can do in a magazine article is give you a brief but interesting and hopefully useful introduction to the subject.
A set of four drawings accompanying this article shows how carbon content affects the crystalline appearance and the physical characteristics of iron and, therefore, also of steel. The percentage of carbon in a specimen has direct effect on strength and other characteristics. One can get some idea of this from the following list:
• Wrought iron Traces to .08%
• Low carbon steel .10 to .30%
• Medium carbon steel .30% to .70%
• High carbon steel 70% to 2.2%
• Cast iron 2.2% to 4.5%
There are hundreds of steel alloys, each one tailored to best suit the requirements of a particular application. Sheet steel for auto body parts has to behave well in the dies that stamp it to shape. Steel for gears has to machine well while at the same time being strong and durable. Steel for cutting tools should be hard but not brittle. For some uses, steel alloys have to be amenable to heat-treating. Manufacturing enterprises in small or undeveloped countries are often handicapped by not having access to the many grades of steel readily available to companies operating in large or highly industrialized countries.
This brings us to the 4130 steel which is used for aircraft tubing as well as many other things. Under the SAE system of classifying steel alloys, the number 4 tells us that it’s a molybdenum-type alloy. The number 1 tells us that it contains 1 percent of chromium., and the 30 tells us that it contains 0.30 percent of carbon. This alloying produces a steel that takes well to welding, is capable of being heat-treated to increase its strength, and which has a fine grain structure that gives good shock resistance. There are small traces of other substances such as sulphur, phosphorus, silicon and copper in the steel as well. In older literature the designation X4130 is often encountered. This is a grade that contains a slightly higher chromium and lower manganese content. The form 4130N means that the steel so identified has been normalized in the factory to relieve stresses caused by manufacturing pro-cesses. Most steel today is normalized, not just the aircraft grade.
Commonly called "chrome-moly" (or chromoly) by aircraft workers, this steel is a strong alloy. The 1020 grade, commonly called "carbon steel" or "mild steel," was used for many early welded steel tubing fuselages. It had a tensile strength of around 65,000 pounds per square inch and was not amenable to heat treating. The 4130 grade has a tensile strength of around 100,000 pounds per square inch, heat-treatable to 180,000 or more pounds. To be on the safe side, many designers do strength calculations on the assumption that the welding process will reduce strength figures by about 20 percent.
Different figures for steel properties are given in different books for such reasons as the facts that strengths can vary with the thicknesses of specimens being tested, the figures given out by different steelmakers can vary, and the processing operations used in manufacturing can also vary. Some of the confusion surrounding the strength of welded steel structures is the result of careless use of technical terms. Few present-day aircraft mechanics are well-informed on the matter of heat treating steel, by reason of working on airplanes that are built largely of aluminum alloy. Another factor is that around 1950 book publishers came to feel that welded steel tubing construction was on its way out so from then on published less and less about it.
It could justifiably be said that the homebuilt aircraft movement that started to grow after 1950 has been obliged to ferret out and relearn a lot of things that have been forgotten by professional aviation engineering and educational people whose task is, naturally, to keep up with the steady flow of new developments. It’s important to remember that each technician is an individual. He or she has studied in different schools, read different books and worked on different things in different places. It’s to be expected that there will be differences of opinion This helps us to avoid frustration and exasperation.
Two months before starting to write this article we wrote to the appropriate FAA office inquiring about the absence of normalizing information in AC 43, but had not received an answer yet as this article went to print. While this is frustrating, we also realize that personnel who worked for the old Bureau of Air Commerce and Civil Aeronautics Administration in the heyday of steel tube construction have long since passed from the scene. In fact, many of the people working for today’s FAA had not even been born when the use of steel tubing reached its high point in the construction of World War II training and utility aircraft. There’s also the fact that FAA engineering people now have to deal with very large, fast and complicated airplanes and can’t find time to deal with what are, to them, peripheral matters.
We asked one major manufacturer of TIG and MIG welding machines how they felt about normalizing 4130 aircraft tubing and were told that their highly trained metallurgists were still working to finalize a company position on the matter. We say these things to help you understand why pat answers to many aircraft construction questions are often hard to get.
It’s also important to remember that much of the technical reading matter now available to Joe Homebuilder was written and published from 50 to 70 years ago. And today, when bringing product liability lawsuits against manufacturers has become such a widespread practice, companies have become very cautious about releasing technical information. This is particularly true where amateur-built airplanes are involved.
The term "heat treating" covers a wide range of procedures and equipment used to tailor the characteristics of many alloys of steel and other metals. And it’s not a process confined just to the aviation field. It often pays off handsomely to look outside the field of aviation literature.
At one public library in our area we found the 1990 edition Encyclopedia Britannica, and its coverage of iron and steel was modest. In another library we found the 1957 edition of this encyclopedia, and it contained an extensive and very informative coverage of the subject.
In yet another library we came upon a recent edition of Manufacturing Processes by Amstead, Ostwald and Begeman, John Wiley & Sons, ISBN 0-471-06245-6 and a 1981 edition of Basic Manufacturing Processes by Kazanas, Baker and Gregor, Gregg Division of McGraw Hill Company, ISBN 0-07-033465-X. They both had great coverages of heat treating.
Ask libraries in your area for an encyclopedia called the McGraw-Hill Encyclopedia of Science and Technology and look up the section on Steel. Suggested reading for the serious researcher is Principles of Heat Treatment of Steel by George Kraus, American Society for Metals, Ohio, 1981 and 1964, ISBN 0-87170-100-6.
TM Technologies, P.O. Box 429, N. San Juan, CA 95960 has recently produced a 4-1/2 hour video that goes into great and authoritative detail about all aspects of aircraft welding.
If you attend AirVenture at Oshkosh, plan to spend some time in the Boeing Aeronautical Library and pore through such old but informative books as Wells’ Manual of Aircraft Materials and Manufacturing Processes, Airplane Welding and Materials by Johnson and Aircraft Materials and Processes by Titterton. Their chapters on heat treating are good.
When airplane factories with exacting government contracts were producing military planes with welded steel fuselages, they put the completed fuselages into large, specially built ovens and heated the entire fuselage frameworks to a suitable red heat. Then the heat was reduced at controlled rates determined by engineers having knowledge of metallurgy. This did a great job of both relieving locked-in stresses created by welding heat and heat treating the steel to develop its full strength.
This process is, of course, totally out of the question for the homebuilder welding up a fuselage in his garage. When general aviation mechanics encounter airplane components stenciled "Heat Treated Member," they are usually landing gear struts, engine mounts and other parts small enough to be put into the kind of more readily available ovens used for many commercial heat treating jobs.
As we previously mentioned, much of the confusion which exists today is the outcome of careless use of technical terms. So let’s now get things straight!
Heat Treating — heating and cooling metal in a certain and carefully controlled way so as to change the physical properties. Depending on the procedure used, steel can be made hard to resist abrasion or soft to facilitate machining it. By means of carefully controlled heat treatment, internal stresses can be removed, grain size reduced, toughness increased or a hard surface produced on a ductile interior. It’s a process to be performed by professionals making use of specialized ovens. It takes time for thick pieces to heat up uniformly, and the rate of cooling has to be carefully chosen and controlled. Heat treating does not exist in the homebuilt airplane field.
Annealing — to heat the metal to a specified temperature, around 1550 degrees F. in the case of steel, allowing it to "soak" for a certain time to assure complete penetration of the heat, and then allowing it to cool SLOWLY. Copper, on the other hand, is annealed by heating it until "peacock" colors show on its surface and then plunging it into cold water. Annealed copper tubing is easier to bend. Steel may be annealed to soften it so that it may be more easily machined or cold-formed.
Much of today’s confusion is the outcome of people using "annealing" and "normalizing" interchangeably.
Normalizing — In a steel mill this means to give new steel a heat treatment that will get rid of internal stresses. In an airplane factory or workshop it means to return steel to this condition, with particular reference to grain size which significantly affects strength after welding. Steel melts at 2600 degrees F. To normalize it after welding, it’s heated as close as possible to 1575 degrees F. and then allowed to cool in still air that’s at room temperature. And by this is meant truly still air! If someone opens a door to come into your workshop while you’re welding, even a mild draft can monkey wrench the situation. Even the cool inert gas flowing from the nozzle of a TIG or MIG torch can put an undesirable draft onto hot metal as the torch is withdrawn from a puddle of weld metal.
Stress Relieving — what the amateur airplane constructor wants to do. It’s vital to understand that the word "normalizing" has been so widely misused to mean stress relieving that sight has all but been forgotten of the fact that after doing welding on 4130 steel, what is really needed is to relieve the stresses put into the affected metal by its having been heated up to the melting point. There’s expansion and then contraction, affecting heated but unmelted metal close to the actual weld puddle. The desired temperature for stress relieving is in the 1150 to 1250 degree range. Bad literature and sloppy shop talk has so confused normalizing and stress relieving as to drive Joe Homebuilder to distraction. "Normalizing" occurs at 1200-1575 degrees and stress relieving at around 1200 degrees.
Some old books say to normalize — meaning, of course, to stress relieve — by heating welded areas until they are cherry red and then allowing them to cool in still air. Others say to heat until it’s blood red. Just how hot are these heats? Steel heated red hot in a gloomy work area will look different than steel heated in a brightly lighted shop. It’s like bad cookbooks that say to add a pinch of this and a dash of that.
We previously mentioned that a lot of literature now available to homebuilders is 50 or more years old. Therefore it doesn’t mention the existence of Tempil heat indicating crayons. These appeared on the scene about 40 years ago and it’s surprising how many airplane builders and mechanics have never heard of them. They are available at welding supply stores.
Tenpil crayons are compounded so that the marks they make on metal melt away when the desired temperature is reached. They are made in 105 different heats ranging from 100 to 2500 degrees. Get to know them and forget all about cherry and blood reds! A metallurgist employed by the maker of Lincoln arc welding equipment suggests using an 800 degree Tempil mark about an inch away from the actual weld area on 4130 tubing. This has to do with how the actual weld metal and surrounding metal are variously affected by heat.
However, there’s more to the matter. Some writers who are emphatic about the impossibility of "normalizing" or stress-relieving welded steel tube structures without resort to a heating furnace point out that it’s anywhere from difficult to impossible to uniformly heat a cluster joint with an oxyacetylene torch. When the flame is directed at one side of the cluster, the tubes that join the cluster on the other side pull heat away from the hot metal and it turns some shade of rather dull red. A few books helpfully suggest using two torches. Of course Joe Homebuilder isn’t keen on buying and then figuring out how to hook up a second torch. Always remember that textbooks are written for students aiming at airplane factory jobs or going for an FAA mechanic license. Homebuilders have to use their heads. We have found that something as simple and inexpensive as a home workshop-type propane torch usefully helps to keep cluster welds uniformly hot.
Some people, in their unfamiliarity with the subtleties of aircraft welding techniques, have completed all the welds in a fuselage and then gone all over the structure again, separately reheating each joint and cluster weld to redness. This wastes time and welding gas. While this won’t please welding perfectionists, from the many things we have read it seems the most reasonable thing for Joe Homebuilder to do is this: After all the welding is done on a particular joint, and while most of the tubing is still fairly hot, pull the torch away from the actual weld so that the bushier part of its flame bathes the weld area as a whole. Keep moving it all around the area to get it as uniformly hot as possible, and draw it away slowly, taking something like a minute or more. This will give you at least enough stress relieving to help your peace of mind. It is very wrong to pull torch flames completely away from weld puddles quickly. The puddle will then cool so abruptly as to "pop" and leave a pinhole. Torch tips used for welding thin aircraft tubing have small orifices. One book suggests that we change to a large tip to get a bigger, bushier flame for stress relieving. But, a weld area will cool down appreciably in the time it takes to change tips.
This is a time when one or more propane torches might prove to be helpful.
Very few A&P mechanics and homebuilders know that 4130 steel is classified as "air hardening" by metallurgists. This means that after having been welded, it does in fact return pretty much to its original state and is thus plenty strong. The misspoken "normalizing" homebuilders do is to relieve locked-in stresses.
While researching this article, we did a lot of reading and corresponded with several welding people. So much has been published about the supposed vital importance of "normalizing" that it comes as a shock to many to learn that a lot of steel tube fuselages have been built without this being performed, and have given long and safe service. For example, we struck up a correspondence with a man who had been an engineer with Piper Aircraft for 20 years and he stated confidently that Piper never "normalized." The reason is partly cost, in terms of both labor and the amount of gas used. We’ve read of engine mounts and seat frames that were also not thus treated and which never suffered as a consequence. To be fair, of course, we have also read of welds that have failed quickly and sometimes inexplicably. These differences apparently involve who did the welding and under what conditions. If you are worried about the fact that steel tube parts in an airplane construction kit have not been "normalized" by the kit maker, ask him to explain why he does not do it. He may have quite logical and acceptable reasons which will banish your worries.
Another person with whom we struck up a most informative correspondence was Chris Fischer, president of Stress Relief Engineering in California. Like most aviation people, we had never before heard of what is called "vibratory stress relief," another outcome of having to rely so much on old aviation literature. Actually, this method has been around for over 35 years and is widely used by a surprising array of manufacturing and contracting firms both in and out of the aviation field.
Because it is interesting and seems likely to show up more and more in our kind of aviation with the passage of time, we’ll tell you something about it. While the subject becomes quite technical as one gets well into it, basically the process involves sending low-frequency, high-amplitude vibrations through a structure. The low-frequency vibrations serve to carry high-amplitude energy to the unit being treated. The heavy vibrations produce a load that is superimposed on the existing stress patterns so as to result in a reduction of peak residual stresses. This produces a more dimensionally stable product and reduces the random distortion that often exists in unstable workpieces. It’s all highly technical, but as far as the customer is concerned, the method works very well and does not require the often tricky and expensive application of heat.
The SRE equipment consists of two components. One is a vibration generator roughly the size of an office typewriter. This is firmly clamped to larger, heavier or immobile structures. Smaller metal objects such as engine blocks, crankshafts and connecting rods, engine mounts, 4130 welded race car frames or smaller airplane fuselages are firmly affixed to the platform of a table-like assembly called the vibration table. Its top is attached to the legs by flexible mounts, and the vibration generator to the underside of this top.
The other component of the SRE system is the castered control unit, which is about the size of a mechanic’s wheeled toolbox. The operator sets the controls to best suit the characteristics of the item being treated and the job is completed in anywhere from several minutes to half an hour. Welded assemblies can be treated either while the welding is being done or afterward.
This SRE equipment is too expensive to be practical for use by individuals and smaller shops. However, at this time a service center is being established in the Midwest to provide this service to interested parties, and in time such centers will appear in other parts of the country. The method has many applications. For example: when a high performance engine heats up, stresses can cause distortions that result in a loss of horsepower or even component failure. Auto racers make much use of SRE. If you are seriously interested in learning more about this method of stress relieving, contact Stress Relief Engineering Co., 1725 Monrovia Ave., Suite A-1, Costa Mesa, CA 92627. Phone 949-642-7820. Fax 949-642-0430.
Because literature published long ago says to do it, today’s homebuilders are very conscientious about oiling the insides of the tubes that make up welded fuselages and other components. So it comes as a shock to them to learn that the fuselages of many popular factory-built lightplanes were never given such treatment. It takes time to run oil in and then drain it out, and production cost has to be closely watched in any highly competitive industry. As all the tubes had closed ends, manufacturers assumed there was no way for water to get inside them. Also, they never thought that their planes would still be in service 50 and more years after manufacture. Every airplane mechanic has found his share of internally rusted tubes in the rear sections of taildragger fuselages and the landing gear areas of nosewheel ships
So after your welding and stress-relieving has been done, if only for peace of mind, you will want to rust-protect the tubing. Long ago a product called Lionoil was marketed by one of the dope and paint firms then active in the aviation field. It appears to have been based on linseed oil to which had been added liquids that made it penetrate small voids readily but also dry at an accepted rate. Present-day aircraft supply firms offer equivalent products.
Old literature mentions linseed oil so often that you should know something about it. It’s made from seeds of the flax plant. In hardware and paint stores you will find both raw and boiled linseed oil. Woodworkers and wooden boatbuilders used to brush raw linseed oil onto their wood because it tended to soak in deeply and minimize checking and water-soaking of the wood. It can take from two to six weeks to dry, which is why Lionoil-type tube protectives contained driers. Boiled oil is used by some painters and by gunsmiths for hand-rubbed finishes on gunstocks because it dries in about a day. We have used boiled oil to re-treat the tubing in old airplanes, and found it to be satisfactory.
We realize that we’ve been discussing a vast and complex subject and that some readers may take exception to one statement or another in this article. We will welcome well-reasoned and informative reader comments.