Dear Fellows,
first of all I guess it should be not too late to wish you all a happy new year!
I am coming from Germany and I must say it is great to read these high qualitative contributions. I would like you to know that we have also a welders forum in Germany - coming from the German Welding Society DVS - but it is definitely not to compare with what you have in the United States and coming from the American Welding Society!
Well, although the query regarding "heat input calculation" has been made approximately 2 years ago, the last reply was stated at December the 23rd last year. Therefore I suppose that it should not be too late for giving a short interpretation of the topic "heat input" from my side. The discussion of what "heat input" in arc welding is, is - at least under German experts - much more topical you ever would suppose. And, so far, it is extremely intricate!
Due to, first I would like to ask "What is Heat?" Expressed in a very simplified way, heat is nothing else than the movement of the smallest particles of a material - what kind ever this is. When we are going to arc weld, we are increasing the amount of particle-movement-energy by transferring the "arc energy" into the base material. Due to the fact of energy conservation - i.e. energy neither can be generated nor destroyed but only being altered in its modifications - the arc energy is being won, by bringing electrical energy to create an accelerated movement of particles of the air (e.g. SMAW) or a gas column (e.g. GMAW or GTAW). By dissociating neutral atoms a/o molecules into (mostly) positive ions and electrons thus generating a plasma we win "kinetic-" or "thermal energy" or simply, we generate an arc. Unfortunately this is not the place and it would go to far to advance further into this very interesting subject matter, because whole books have been written to discover the secrets of what really occurs when an arc is being ignited on a work piece and transfers its energy into the material. Only one little detail I would like to mention, normally, when someone is talking about calculating the "heat input" in arc welding, he must also consider that not the entire amount of electrical energy is being transferred as "heat" to the work piece or material, respectively. There are losses of energy being caused by physical influences just like convection (heat flow in gases and fluids), heat radiation (from arc rays or melting bead) which are, however, negligible both in appropriate equations. Or by resistive heating (Joule's heating) of the electrode material. The sum of all these physically based losses must be subtracted from the amount of electrical power (I * E = Watt or J/s) being delivered from the welding power supply. Therefore the final amount of what value of "heat input" is being transferred into the base material by the arc, is lower than the amount of electrical power (I * E) and is being expressed by the so called factor of "arc efficiency" (normally written as the Greek letter "eta" suffixing an "a"). If the expression of "heat input" (whatever it means) should be calculated correctly, it should take the form of: H = eta a * I * E * 60 / s. However, this fact should not change the general conditions for the in 2004 asked question, due to there are also arc welding processes (e.g. Submerged Arc Welding), working with an "arc efficiency" of ~ 1 (i.e. nearly 100% of electrical energy is being altered into arc energy). The basis of the query was - as far as I understood - whether an oscillated melting bead transfers more heat input into the base material than a stringer bead; and thus if the commonly known equation for calculating the "heat input" namely H = I * E * 60 / s, is usable for this venture. Alright, let us forget all the necessary details of solid state- and plasma physics being normally crucial for calculating the sequences in arc energy- or heat transfer into the base material. For this simplified approach for finding out if there is being transferred more heat input into the base material when the welding is being executed with an oscillating motion, let me make a short description of a consideration about the material, being molten off from the electrode per unit time. As far as I have seen over the course of discussing "Bonniweldor's" question, all fellows are wonderful right in interpreting what is meant, and thus, in making their single predications in regard to the topic. Therefore I have hesitated on myself if it would be actually necessary to write something additional to this topic. But I will try it and thus I would like you to follow me when I am going to express the inseparably interconnection between "heat input" and "heat flow" by using the subsequent example...
Imagine you are going to use Manual Shielded Metal Arc Welding. Presume further you have to fill a Single V-Groove cross section and simplify, that the depth of fusion - although you are going to weld manually - is constant under every circumstance. Also every other condition, coming from the base materials and environmental side, just like materials composition, heat conductivity, heat capacity, base materials temperature prior to welding or spatter loss is equal. Now consider that the voltage and the amperage you are going to use for filling the V-Groove cross section, are ensured to be constant over the entire welding procedure. Likewise it should be implied, that your skill for performing the manual welds, is just like "robot comparable". Furthermore physical influencing factors like heat convection, radiation and resistive heating of the electrode can be neglected for simplifying the whole assumption. Now presume you are going to need a definite number of electrodes with a definite diameter (neglect the covering of the rods) for filling the cross section. Using a defined amount of filler metal means, melting off a defined amount of weld metal and thus, achieving a defined melting rate - always under presuming all other peripheral conditions being held constant. And now imagine that you will have the opportunity to choose the way of filling the cross section - stringer or weaving the layers into the groove! The only stringent condition is, that you have to perform the cross section filling operation under using the constant amount of electrodes, either you are weave- or stringer beading, and thus the melting rate is constant in both. My question to you all dear fellows: "Would you agree that the total amount of arc energy, generated and being transferred into the base material while melting off the constant number of electrodes for filling up the defined cross section of the V-Groove, is equal, indifferently whether you are using stringer bead or weaving bead technique?" Well, when you are going to reply "Yes", I would be on your side, however, strongest simplification of all the very complex physically interacting factors is assumed. But! This, what we have tried to carry out was the calculation of "thermal energy input" by using the "arc on time" by presuming a defined amount of filler metal, being molten off under using defined electrical power conditions (I * E) for filling a defined cross section. Therefore - under theoretical considerations - I can not fully agree with "DGXL" when he says that a 6"-weld, using stringer bead technique, takes less time than a 6"-weld performed under using weaving bead technique. This difference in welding time may be pure reality in daily practice and "DGXL's" predication must be confirmed by using own practical experience, no doubt, but it should not count when we are talking about calculating the value for thermal energy input from the pure theoretical standpoint. Nevertheless - when I am allowed to say - "DGXL" is right when he explains that the difference in welding time has a deep impact on the "heat input" (per unit length). Since the time for performing the weld under using weave beading technique is increased, in opposite the welding velocity is theoretically decreased. Therefore the energy- or lets say "heat input" into the base material is also increased for the case of weave beading technique. What are we now going to do when we have to calculate the theoretical "heat input" under using the expression "Bonniweldor" has stated at the beginning of the entire forum discussion (H = I * E * 60 / s)? Here I would like to suggest to follow the colleagues who have described the meaning of "arc on time". Once again it should be presumed that a weld had to be carried out by using the MSMAW-process. And weaving technique should be the condition. Then it should be possible to measure the time needed to melt off 1 (one) electrode in stringer bead technique and thus to create a proper length of the weld bead ratio (accurate relation of depth of fusion[h] to width of fusion[w] = [h]/[w]). Thereby it might be possible to use the expression H = I * E * 60 / s, to calculate the "heat input" - with simplification of neglecting the factor of arc efficiency (eta a). Then afterwards, when performing the weaving bead technique to carry out the weld, one can multiply this value of theoretical "heat input" by the number of electrodes been molten off for the entire operation. Hereby to achieve the final value of theoretical "heat input". So far to the term "heat input" what means from my point of view, the theoretical calculated value of thermal energy being won by transferring electrical energy to kinetic- or thermal power and coupling shares of it into the base material.
Coming now to the predication posted by "Phil Thomas" on Dec. 22nd, where he assumed, that we are going to have two parallel discussions - even "heat input energy" and "heat flow". I agree with him. From my point of view, much more important than the theoretical "heat input", calculated by the expression above, is the impact of thermal energy to the base material properties. And thus, I would like to come in a very short way to this topic being so important and invariably coupled to the theoretical "heat input" calculated by H = I * E * 60 / s. As you know any kind of point energy source - just like an arc - which is moving with a defined velocity over a defined base material, creates specific lines or areas, respectively, of constant heat content, so called "Isotherms". It is a great difference between only the theoretical "heat input" (per unit length) and the extent of the isotherms being created by the heat. And thus it is a great difference if you are going to weave or stringer the bead! Why..? Well, since now, when we are going to deal with the impact of the transferred heat to the material, it is no more possible to neglect specific geometrical and physical base material properties like thickness, heat conductivity, heat diffusivity, density etc. and, in some cases, similar properties for the filler material. Basically the thermal conductivity of metals is a temperature depending property, i.e. the higher the temperature of the material is (up to the specific boiling point - then the conductivity is rising again) the lower is its ability to conduct heat - mentioned by the way, consider also the higher electrical resistance of most metals at higher temperatures, since heat- and electrical conductivity are based on the same physical coherences. They correlate both. Let us presume once again to use the Manual Shielded-Metal-Arc-Welding and carrying out the stringer bead technique. This is, dependently to the electrical power (P = I * E = J/s), the welding velocity (s in cm/min.) and neglecting the arc efficiency, calculable by H = I * E * 60 / s. Correlating to this heat- or energy input and depending to the material properties, specific limited isothermal areas being generated around the arc and its way over the material. Corresponding to these coherences - particularly for low alloyed or high strength steels - the crucial important cooling time interval between 800°C and 500°C is being influenced. Within this interval the thermodynamically sequences are running in a way, being deciding for the later materials microstructure. The higher the thermal energy per unit area per unit time, the larger the isotherms, the lower the cooling rate and thus the longer the cooling interval between 800°C and 500°C. The theoretical instantaneous cooling rates again, are depending basically on:
- Geometry and dimensions of the work piece
- Effective heat input per unit length
- Height of preheating temperature
Instantaneous cooling rates can be calculated but, I would like to forbear on these calculations here due to their relative intricacy. What is important - and this has also already been expressed by "reddoggoose" on Sept. 08th, by predicating: "...as such you can see that a lower travel speed number in the denominator of the equations gives an answer of higher value..." - is, that for changing the conditions for cooling times and -rates, either the electrical power must be increased, or the welding velocity must be decreased. Or, to meet the third aspect in the list above, preheat the component to be welded. So far so good. In case of stringer beading, the material is being molten in a small area by the arc. Additional the energy source is being moved onwardly with a defined velocity. Depending to the materials dimensions, geometry and physical properties, the isotherms are specific in extent and thus the instantaneous cooling rate or t 8/5 is specific to the isothermal areas, respectively.
And now it comes. Differently to stringer beading technique in case of weave beading, the situation is comparable to the physical circumstance of so called "multilayer welding with short seam lengths". In this case the temperature of the area adjacent to the seams being welded, is increased, by welding short time sequenced several seams. For better understanding, presume a V-Groove weld with a bevel angle of 30° on a plate being 15 mm thick, 200 mm wide but only 70 mm long. Now you have to fill the cross section. The time needed for welding all seams, necessary to fill the entire cross section, may be shorter, than if the plate would be 700 mm long. Thus, the time interval every single preceded seam is being welded over by the following one, is shorter than being it in the latter case (700 mm long) and thus the cooling intervals become likewise shorter. Therefore it comes to a kind of "compression" of temperature which increases, with the amount of layers being welded over the preceded ones. The total height of work piece temperature increases and the cooling rate - also between 800°C and 500°C - decreases, until the final layer is being welded and the work piece cools down from the achieved common temperature level. Comparable with this is the weave beading technique. By oscillating the melting bead, the preceded "layer" - long, as wide as the total width of oscillation - is being welded over in a very short interval of time, therefore the total amount of thermal energy (temperature) per unit time per unit area does rise and thus the cooling rate does decrease. Mostly the reason for an increased grain coarsening and thus, deteriorated mechanical properties of the welding joint. This is one of the reasons for the recommendation to use the stringer bead technique for welding e.g. High Strength Low Alloyed Steels. Determining the length of a weld under using defined electrical parameters in relation to base- and filler material properties is the way to combine both "heat input" just as asked by "Bonniweldor", and "heat flow", as predicated by "Eutectic". The equations posted by "Eutectic" - allow me to congratulate him for the excellent explanations - dealing with 2- and 3-dimensional heat flow calculation. These are the base for engineers working in the field of crane-building and -construction. Those people are only looking on the - as we in Germany say - "t 8/5" for predicting the materials- or joint properties, respectively. No one wants to know there, what the theoretical "heat input" calculated by the expression H = I * E * 60 / s, is in fact by its own. Only the reasonable combination of "heat input" and "heat flow" and their interactions are of crucial interest. Rosenthal's equations were as far as I know the one of the first groundbreaking attempts to calculate the energy input of a moving source, i.e. an arc, and deliver also today relatively good predictions in some basically questions of heat transfer processes like arc welding, although intermediately they have been improved in many different ways. I guess this is what "Lawrence" on Sept. 27th 2004 meant ("...it will smoke your hard drive."). Finally please allow to let me try to reply the question posted by "Cain" on Sept. 27th 2004. Yes, the heat transfer efficiency is dependent to, or based on the welding process, respectively. Particularly the deeper trials of calculating heat input in case of using modern arc welding processes (CMT, STT, AC-MIG etc.) are increased in intricacy. Due to the difficulty to define the correct heat transfer efficiency factors to define the correct way of calculation.
Concluding I would like to say that - from my point of view - mostly it might be a great difference between the values basing on theoretical calculations and those ones, being measured via experimental procedure, e.g. by using calorimetry for defining the net heat input. Since I can imagine that there is also a difference between the theoretical heat input calculated for weaving and the final practical net heat input using this technique (see also "DGXL's" reply on Sept. 25th 2004), it is assumable that in case of practical weaving operation the net heat input might be higher than in stringer bead technique - although the theory would speak another speech.
Last but not least, and before finishing my accomplishments, I am proud to be a fellow of the great American Welding Society and thus having the chance of discussing with you!
Thank you for reading and best regards from Germany,
Stephan