Where the area comes from in each preparation
Every groove weld is the same three pieces of area added together: the bevelled space, the root gap running through the full thickness, and the reinforcement crown standing above the plate. The preparations differ only in the shape of the first one.
A single-V bevels both edges from the same side. With a root face of f on plate of thickness t, the vee is a triangle of depth t minus f, and its area is that depth squared times the tangent of half the included angle. A 60 degree included angle means each edge is cut at 30 degrees from vertical, and the tangent of 30 is about 0.577.
A double-V does the same from both sides, so each vee is half as deep. Because the area goes as depth squared, halving the depth quarters each vee, and there are two of them, so the total is exactly half the single-V. That factor of two is the entire argument for double-sided preparation and it is worth stating precisely: half the weld metal, roughly half the arc time, and roughly half the consumables, on the same plate at the same angle.
A single bevel prepares one edge only and leaves the other square, which makes a right triangle rather than a symmetric vee. Here the angle input is the angle of that one bevelled edge rather than an included angle between two, so the area is depth squared times the tangent of the full angle, over two. That is the preparation for a T-joint that needs full penetration, and for a butt where one plate cannot be turned or cut.
A square butt has no bevel at all and its entire cross-section is the root gap through the thickness. It works on thin material where the arc can reach the far side, and it stops working as soon as the plate is thick enough that it cannot.
The root gap is not a small thing
People treat the root gap as a fit-up tolerance and it is not, it is a design decision that adds area through the full thickness of the joint. On three quarter inch plate, an eighth inch root gap contributes 0.094 square inches, which on a 60 degree single-V is roughly a third of the total fill. Opening the gap from an eighth to three sixteenths on that joint adds another 0.047 square inches, which is more metal than a whole fill pass.
It runs the other way too. A tight root gap on a joint that needs one gives an arc nowhere to reach and produces a root that has not fused, which is the defect nobody sees and everybody fears. The gap is chosen for the process and the position, and it is one of the reasons a joint detail is a specification rather than a preference.
Reading the comparison table
The table shows all five preparations at your thickness, angle and gap, so the trade-offs are visible rather than argued. Three columns matter in combination.
| Column | What to do with it |
|---|---|
| Cross-section | Drives everything downstream: fill time, filler, gas, and distortion. It is the number to minimise if you can. |
| Passes | A planning estimate. It tracks cross-section, but interpass handling and cooling scale with it too, so it is the better proxy for elapsed time. |
| Sides | The constraint that overrides the other two. A double-sided prep needs access to both faces, the ability to turn the part, and a back gouging step. If any of those is missing, the preparation is not available regardless of how good its numbers are. |
The last column shows each preparation as a multiple of the smallest, which makes the comparison at a glance rather than by subtraction. On thick plate the spread between the best and worst preparation is often more than two to one, and that ratio applies to filler cost, gas, arc time and distortion simultaneously.
Pass count, and what it is honestly worth
The pass count here divides the fill area by a nominal pass cross-section after setting one pass aside for the root. The default of 0.06 square inches per fill pass is in the range of what a manual welder deposits in a controllable bead in the flat position, and it is smaller out of position because gravity takes the puddle. Change it to whatever your own work produces.
What the estimate cannot represent is that real joints are not filled with identical passes. The root pass has different requirements from everything above it, the fill passes may be split side to side across the width of the groove rather than laid one on top of another, and the cap is sized for appearance and for how it blends into the plate rather than for volume. A joint that this page says takes eight passes might genuinely be welded in six or in eleven depending on who is welding it and how.
Use it for planning: how long a joint will take, how much wire to have on hand, whether a job is a morning or a day. Do not use it to tell a welder how to fill a groove.
What each preparation costs to make
The metal cut away to create the bevel is reported per foot, and it is the part of the comparison people forget. A double-V removes the same total volume as a single-V from the plate edges, but it does it from both sides, which means two setups on a bevelling machine, or two passes with a grinder, or a plasma cut that has to be flipped. Preparing an edge is real work and it is not free just because it happens before the welding starts.
Thermal cutting the bevel changes the edge as well as the shape. A plasma or oxyfuel cut leaves a heat affected edge and, on some materials, a hardened layer that has to come off before welding. That is another reason the choice of preparation is a whole-process decision rather than an arithmetic one. For the cost of the cutting itself, the plasma cutting cost calculator handles consumables and cut time.
Where this connects
The cross-section this page produces is the input to three other calculations. The weld cost calculator turns it into filler, gas and labour cost. The travel speed calculator turns it into a travel speed and arc time for a given wire feed. The distortion calculator turns it into the movement you should expect, which is where the double-V argument gets its second and larger benefit: balanced about the mid-thickness, the angular pull very nearly cancels.
What this page will not do
It will not tell you how big the weld should be. Weld size for anything that carries load, restrains load, or lifts anything at all is an engineering specification, produced by someone qualified to produce it, and executed by a welder qualified for that joint in that position. Inspection of the result is a third discipline again, with its own qualifications. Lifting points, trailer hitches, vehicle suspension and structural connections are all in that group, and a calculator that reported a weld capacity for them would be doing real harm.
So this page takes a weld size you have already been given, or that you have chosen for a non-critical job, and works out what follows from it. If nobody has given you a size and the failure of the joint would hurt somebody or cost a lot of money, that is the signal to stop calculating and get the specification.
The hazards of welding itself are set out in full on the travel speed calculator, and joint preparation adds cutting and grinding to that list without changing any of it.
Questions people ask
Why is a double-V exactly half the fill of a single-V?
Because the area of a triangular groove goes as the square of its depth. Splitting the preparation into two vees from opposite sides halves the depth of each one, which quarters its area, and there are two of them, so the total is half. It is exact for the bevelled part of the joint. The root gap contributes the same area to both preparations because it runs through the full thickness either way, and the reinforcement cap is larger on a double-V because there are two faces to crown, so the totals are close to half rather than exactly half. The table shows both the groove volume and the total so you can see which is which.
When is a double-sided preparation not worth it?
When you cannot turn the part, when you cannot reach the far side, or when the back gouging step costs more than the metal it saves. Back gouging is a real operation with its own equipment and its own mess, and on a light plate where the fill is small anyway, the saving does not repay it. It also stops being available the moment a joint is welded into a larger assembly, which is why a lot of shop-fabricated work uses double-sided preps and a lot of site work does not. The arithmetic favours double-sided almost always; access and sequence are what decide it.
What included angle should I use?
That is a specification question rather than a calculation, and it is decided by the process, the position, the thickness and what the joint has to do. What the arithmetic shows is the shape of the trade-off: the fill area goes as the tangent of the half angle, so opening a 60 degree groove to 75 degrees adds about 32 percent more weld metal, and closing it to 45 degrees removes about 28 percent. Narrow grooves save a lot of metal and give the arc less room to reach the sidewalls, which is where sidewall fusion defects come from. That balance is exactly the kind of thing a written procedure exists to settle, and not the kind of thing to settle from a table on the internet.
Does the reinforcement cap need to be there at all?
It is a normal outcome of filling a groove rather than a target, and it is included here because it is real weld metal that you pay for and that adds to the shrinkage. The area used is an approximation of a shallow parabolic crown, two thirds of the width times the height, which is close enough for estimating. Excess reinforcement is not a virtue: a tall crown adds metal, adds distortion, and creates a sharper transition at the weld toe. Whether it is acceptable, required or has to be dressed flush is a specification matter that depends entirely on what the joint is for.
Can I use this for a fillet weld?
No, and the distinction matters. A fillet weld sits in the corner of two members and has no groove at all, so its cross-section is the leg squared over two rather than anything on this page. The travel speed calculator handles fillets directly, and the weld cost calculator handles both. This page is specifically about what happens when the edges of the material have been prepared to let the weld reach through the thickness, which is the situation where the preparation choice is worth a comparison in the first place.