Shoulder Screw Calculator

Q: Are metric shoulder screws available?

Yes. The standard is ASME B18.3.3M, Hexagon Socket Head Shoulder Screws (Metric Series), which covers 6.5 to 25 mm nominal shoulder diameters. It was published in 1986 and reaffirmed in 2008, and ASME now lists it under stabilized maintenance rather than as an active standard, which is why the usual answer to this question is wrongly no. Distributors stock the parts, and they follow the same rule as the inch series: a 6 mm shoulder on an M5 thread, 20 mm on M16, 24 mm on M20. The table on this page is inch only because the inch figures could be confirmed in three independent sources; the metric ones could not, so they are not printed.

Every ASME B18.3 dimension for all fourteen hex socket head shoulder screw sizes, in inches and millimetres — with the one subtraction that decides whether the joint actually turns made the headline, and the shoulder checked as the journal it is.

Last updated 24 Sep 2026

Shoulder screws, ASME B18.3

Size and grip → the dimensions and the clearance

Shoulder screw size, ASME B18.3Fourteen sizes, 1/8 to 2 inch shoulder. Read the thread beside each one: it is always smaller than the shoulder, and the pairing is the standard’s rather than a supplier’s choice. There is no 1/4 inch shoulder screw with a 1/4 inch thread, and there never was.

The lengths you are about to type are inThe part is an inch part, so inches is the default. Every dimension the page prints comes out in both.

Shoulder LENGTH of the screw, as boughtThe ground, unthreaded length between the underside of the head and the start of the thread. It is the dimension the part is ordered by, and it is NOT the overall length.

GRIP: total thickness the shoulder passes throughEverything the shoulder goes through and nothing it does not: the rotating part, plus any washer or spacer under the head, and not the plate the thread screws into. This subtraction is the one people get wrong.

Radial load carried by the shoulder (N)Newtons, whichever length unit you chose above. Used for the bearing pressure, the shear stress and the bending stress lower down. Leave it alone if you only came for the dimensions.

Not a circuit: a section through the joint, drawn to the proportions ASME B18.3 actually specifies for the 1/2 inch size at a scale of one inch to 150 units. Read it from the top down and the part explains itself. The head is one and a half times the shoulder across, which is why the counterbore is always larger than the running bore. The shoulder is a plain ground cylinder and the rotating part turns on it. The thread is visibly narrower — one standard size down, 3/8-16 under a 1/2 inch shoulder — and it screws into the lower plate. The gap arrowed at the top right is the whole point: the shoulder bottoms on the seated face BEFORE the head reaches the rotating part, so the screw can be taken to full torque while the part stays free. It is drawn exaggerated here; in a real joint it is a few thousandths of an inch. Take it away and the joint is a bolted one with an undesigned clamp load, which is exactly what you get if you substitute a cap screw, because a cap screw has no shoulder to bottom against at all. The bar at the bottom is live: it is your own clearance on a nought to one millimetre scale with the working band marked at 0.025 and 0.5 mm, and it reads empty rather than negative when the subtraction has gone the wrong way.

0.254mmExample

A 1/2 inch shoulder screw with a 1.000 inch shoulder, through a 0.990 inch plate, carrying 4 kN across the shoulder

One subtraction, and then the shoulder treated as what it is: a short journal with a screw on the end

running clearance = L_shoulder − grip · p_bearing = P / (d_s · t_grip) · τ = 4P / (π d_s²) · σ_b = 16 P t_grip / (π d_s³)

L_shoulder: the ground, unthreaded length between the underside of the head and the first thread. It is what the part is ordered by and it is not the overall length, which also includes the head and the thread
grip: everything the SHOULDER passes through: the rotating or sliding part, plus any washer or spacer under the head. Not the plate the thread goes into. Getting this wrong by including the tapped plate is how a joint ends up with a negative clearance on paper and a clamped pivot in the machine
d_s: the shoulder diameter, which the standard holds at the nominal size MINUS 0.002 inch at every one of the fourteen sizes. That is a rule, not a table, and it is the clearance strategy: the shoulder is made undersize so the bore can be made on size
p_bearing: the projected bearing pressure, load divided by diameter times bearing length — the standard journal convention, so it is a nominal pressure on the projected rectangle and not a peak contact stress. The allowable belongs to the BUSHING or the bore material, not to the screw, so no limit is printed here
σ_b: bending stress in the shoulder, taking it as a cantilever built in at the seated face with the load acting at mid-grip, so the lever arm is half the grip. That is the honest model for a stripper bolt or a single-sided pivot and it is conservative for a load spread over the full bearing length. A clevis that loads the shoulder in double shear is a different and much kinder case, and the page prints that shear figure too
A = 0.7854(D − 0.9743/n)²: the unified-thread tensile stress area, D the major diameter and n the threads per inch. It reproduces the published figure at all fourteen threads in the standard. Worth knowing: this is NOT the same definition ISO uses for a metric thread, which takes the mean of the pitch and minor diameters; the ISO form gives 0.9 to 2.1 per cent more area for the same thread, so an inch and a metric stress area are not quite comparable quantities

Worked example

A 1/2 inch shoulder screw with a 1.000 inch shoulder, through a 0.990 inch plate, carrying 4 kN across the shoulder
THE SUBTRACTION FIRST, because it decides whether the joint is a pivot at all. The shoulder is 1.000 in and the grip is 0.990 in, so the shoulder stands 0.010 in proud of the stack — 0.254 mm, ten thousandths. The shoulder bottoms on the tapped face, the screw can be taken to full torque, and the plate is left with 0.254 mm of axial float and free to turn. Reverse those two numbers and the head reaches the plate first: the screw then clamps it, and no amount of torque will make it a pivot
THE SHOULDER IS NOT A HALF INCH. ASME B18.3 holds it at the nominal minus 0.002 in, so the maximum is 0.498 in = 12.649 mm. Ream the bore to a nominal half inch and you have given away two thou of diametral clearance before adding any running allowance at all. That figure is in the standard and is missing from most charts
AND THE THREAD IS ONE SIZE DOWN: 3/8-16 UNC, a major diameter of 9.525 mm against the shoulder's 12.649. The step is 3.124 mm, the diameter ratio is 1.328, and the AREA ratio is its square, 1.764. The shoulder's full cross-section is 125.67 mm² against the thread's 49.99 mm² of tensile stress area — 2.51 times as much. The shoulder is the structural member here and the thread is a retainer
AS A BEARING. The projected area is diameter times bearing length, 12.649 × 25.146 = 318.1 mm², so 4 kN gives a nominal bearing pressure of 12.58 MPa. That is low, and whether it is acceptable is a question about the BORE, not about the screw: a bronze bush, a steel bore and a dry plastic bush have allowables an order of magnitude apart, so this page prints the pressure and refuses to print a limit
AS A PIN. In single shear the stress is 4000 / 125.67 = 31.83 MPa; in double shear half that. Bending is the one that bites. Treat the shoulder as built in at the seated face with the load at mid-grip: the arm is 12.573 mm, the moment is 50,292 N·mm, the section modulus is πd³/32 = 198.70 mm³, and the bending stress is 253.1 MPa — eight times the shear. Combined with the direct shear the von Mises stress is 259.0 MPa. Against alloy steel at 140 ksi minimum tensile, which is what AFT quotes for this part, that is comfortable; the lesson is which number was large
THE HOLES. The head is 0.750 in = 19.05 mm across and 0.312 in = 7.92 mm tall, so a counterbore of about 19.84 mm diameter (adding a thirty-second on diameter, which is shop practice and not in the standard) by 7.92 mm deep brings it flush. The thread length is 0.625 in = 15.88 mm, so tap at least 15.88 plus two pitches, about 19.05 mm, so the thread stays off the bottom of the hole and the SHOULDER is what lands. That two-pitch allowance is working practice rather than anything ASME B18.3 specifies, and this page labels it as such

ASME B18.3 hex socket head shoulder screws, all fourteen sizes

ASME B18.3 is cited by number and not reproduced; the printed figures here are Fuller Fasteners’ published table for it, and they were checked against two further sources rather than taken on trust. AFT Fasteners’ ASME B18.3-2003 table confirms every column including thread length for 1/4 through 3/4 inch; EngineersEdge confirms shoulder, head, socket and thread for 1/4 through 2 inch; TorqBolt confirms head, socket, thread and TPI for the three smallest sizes, which neither of the other two carries. Two things to notice. First, the shoulder column is NOT the nominal size: it is the nominal minus exactly 0.002 inch at every one of the fourteen rows, which is a rule rather than fourteen numbers, and it is the whole clearance strategy for a bearing fit — the shoulder is deliberately undersize so the bore can be made to nominal. Second, that same column is labelled differently by different publishers: Fuller, AFT and EngineersEdge print these numbers as the MAXIMUM, TorqBolt prints the three smallest as the MINIMUM of a window whose top is the nominal. The numbers agree and the label does not, so if you are inspecting to a tolerance rather than designing a hole, read the standard. The thread stress area is computed here from the unified-thread expression A = 0.7854·(D − 0.9743/n)² and agrees with the published figure at all fourteen threads; the 1/8 inch thread length is Fuller’s 0.156, against TorqBolt’s 0.157. These dimensions come from a published standard’s table, not from a formula. The standard itself is cited below and the printed values are attributed to the catalogue they were taken from; a different publisher may round differently in the last digit. The same designation can mean different dimensions in different standards families — ANSI against ISO, inch against metric, one national standard against another. The family used here is named beside every figure; check which one your part was made to.
Nominal shoulder	Shoulder dia, max (in)	(mm)	Head dia, max (in)	Head height, max (in)	Hex socket (in)	Thread	TPI	Thread length (in)	Thread stress area (sq in)
1/8 in	0.1230	3.124	0.250	0.125	0.078	#4-40	40	0.156	0.00603
5/32 in	0.1540	3.912	0.281	0.125	0.094	#6-32	32	0.187	0.00909
3/16 in	0.1850	4.699	0.312	0.156	0.094	#8-32	32	0.189	0.01401
1/4 in	0.2480	6.299	0.375	0.188	0.125	#10-24	24	0.375	0.01753
5/16 in	0.3105	7.887	0.438	0.219	0.156	1/4-20	20	0.438	0.03182
3/8 in	0.3730	9.474	0.562	0.250	0.188	5/16-18	18	0.500	0.05243
1/2 in	0.4980	12.649	0.750	0.312	0.250	3/8-16	16	0.625	0.07749
5/8 in	0.6230	15.824	0.875	0.375	0.312	1/2-13	13	0.750	0.14190
3/4 in	0.7480	18.999	1.000	0.500	0.375	5/8-11	11	0.875	0.22600
1 in	0.9980	25.349	1.312	0.625	0.500	3/4-10	10	1.000	0.33446
1-1/4 in	1.2480	31.699	1.750	0.750	0.625	7/8-9	9	1.125	0.46173
1-1/2 in	1.4980	38.049	2.125	1.000	0.875	1-1/8-7	7	1.500	0.76328
1-3/4 in	1.7480	44.399	2.375	1.125	1.000	1-1/4-7	7	1.750	0.96911
2 in	1.9980	50.749	2.750	1.250	1.250	1-1/2-6	6	2.000	1.40525

What the shoulder actually buys you

Every column here is computed, from the shoulder and thread diameters in the table above. The last one is the argument for the part: the shoulder’s cross-section is between two and not quite three times the thread’s tensile stress area at every size — 1.92 at the 1/8 inch size, 2.76 at 1-1/4 inch — so in a pivot or a stripper the shoulder is the strong member and the thread is only holding the screw in. That is why a shoulder screw can be run as a pin without a nut, and why the failure you should expect if you overload one is a sheared or bent SHOULDER rather than a stripped thread — the opposite way round from a cap screw. The area ratio column is the one to read before designing a small pivot: at 1/8, 5/32 and 3/16 inch the shoulder is only 1.10 to 1.13 times the thread diameter, so it buys a fifth to a quarter more area; from 1/4 inch upward it buys between forty per cent and double. The step is not a scaled constant and nothing in the standard says it should be. These dimensions come from a published standard’s table, not from a formula. The standard itself is cited below and the printed values are attributed to the catalogue they were taken from; a different publisher may round differently in the last digit.
Nominal	Shoulder dia (mm)	Thread major dia (mm)	Step (mm)	Shoulder / thread	Area ratio	Shoulder shear area (sq mm)	Thread tensile area (sq mm)	Shear / tensile
1/8 in	3.124	2.845	0.279	1.098	1.206	7.7	3.89	1.97
5/32 in	3.912	3.505	0.406	1.116	1.245	12.0	5.86	2.05
3/16 in	4.699	4.166	0.533	1.128	1.272	17.3	9.04	1.92
1/4 in	6.299	4.826	1.473	1.305	1.704	31.2	11.31	2.76
5/16 in	7.887	6.350	1.537	1.242	1.543	48.9	20.53	2.38
3/8 in	9.474	7.938	1.537	1.194	1.425	70.5	33.83	2.08
1/2 in	12.649	9.525	3.124	1.328	1.764	125.7	49.99	2.51
5/8 in	15.824	12.700	3.124	1.246	1.553	196.7	91.55	2.15
3/4 in	18.999	15.875	3.124	1.197	1.432	283.5	145.81	1.94
1 in	25.349	19.050	6.299	1.331	1.771	504.7	215.78	2.34
1-1/4 in	31.699	22.225	9.474	1.426	2.034	789.2	297.89	2.65
1-1/2 in	38.049	28.575	9.474	1.332	1.773	1,137.1	492.43	2.31
1-3/4 in	44.399	31.750	12.649	1.398	1.956	1,548.2	625.23	2.48
2 in	50.749	38.100	12.649	1.332	1.774	2,022.8	906.61	2.23

Metric shoulder screws, which do exist

The usual thing said about this part is that it is inch only. That is true of ASME B18.3, which is an inch-series document, and false of the family: ASME B18.3.3M, *Hexagon Socket Head Shoulder Screws (Metric Series)*, 1986 and reaffirmed 2008, covers nominal shoulder diameters from 6.5 to 25 mm. ASME’s own catalogue entry for it adds the part that explains why nobody mentions it: the standard is no longer ASME-approved and is retained under stabilized maintenance for historical reference. It is still what metric shoulder screws are sold against, and the three rows above are pairings a distributor lists as meeting it. Note that they obey exactly the same one-size-down rule as the inch series — a 20 mm shoulder on an M16 thread, a 24 mm shoulder on an M20 — and note the first row, which is outside the range ASME states for the standard it is listed under. That is printed rather than tidied away, because a 6 mm shoulder screw is a real thing you can buy and the discrepancy is the distributor’s, not the reader’s. The dimension table on this page is inch only because that is what could be verified in three independent sources; treat the absence of a metric table here as a limit of this page and not of the part. The same designation can mean different dimensions in different standards families — ANSI against ISO, inch against metric, one national standard against another. The family used here is named beside every figure; check which one your part was made to.
Nominal shoulder (mm)	Thread	Within the 6.5 to 25 mm range ASME states for B18.3.3M?
6.0	M5 x 0.8	no, it sits below the bottom of the stated range
20.0	M16 x 2	yes
24.0	M20 x 2.5	yes

Four ways a shoulder screw joint goes wrong, and the subtraction that catches all of them

The first row is the commonest and the cheapest to avoid. Shoulder length minus grip: positive and small is a working pivot, zero or negative is a clamped joint pretending to be one. The third row is worth a second look, because it fails quietly rather than obviously: the joint turns, it just turns with more play than the drawing allowed, and the cause is two thousandths of an inch that are in the standard and not on most charts. This page sizes a part; it does not certify one. Where the answer carries a consequence — a load path, a lifting duty, a pressure boundary, a fastener holding something that can fall — confirm it against the design code that governs the application, and against the manufacturer’s own rating, before relying on it.
What was done	What happens	What to do instead
The shoulder is SHORTER than the stack it passes through	The head lands on the rotating part before the shoulder bottoms out, so tightening clamps the part. It will not turn, and the torque you applied is now carried by a friction joint you did not design	Buy a longer shoulder, or shim the stack down. The page’s headline is exactly this difference and goes negative when you are in this case
A socket head cap screw was used instead	There is nothing to bottom on at all. The clamp load is whatever torque was applied, the thread carries it, and the part either seizes or works loose as the plain shank wears into the bore	Use the shoulder screw. If the part really is meant to be clamped, use a cap screw and say so — but then it is not a pivot
The bore was reamed to the shoulder’s nominal size	The shoulder is 0.002 inch under nominal, so a nominal bore gives two thou of diametral clearance before any running allowance. On a locating pivot that is slop you will feel	Size the bore from the shoulder’s actual maximum, which this page prints, and add the running allowance the fit calls for rather than the nominal-size difference
The thread was taken as the load-carrying member	It is not. The shoulder’s cross-section is roughly two to three times the thread’s stress area, so the shoulder governs in shear and bending and the thread only retains the screw	Check the shoulder for shear and for bending about the seated face, which this page does, and size the tapped depth for retention rather than for strength

A journal with a screw on the end: the two-thou undersize shoulder, the one-size-down thread, and the subtraction that decides whether it turns

A shoulder screw is a journal with a screw on the end of it, and everything about the part follows from that. The ground, unthreaded shoulder is the bearing surface; the thread is one standard size smaller and does nothing but hold the screw in. ASME B18.3 pairs each of fourteen shoulder diameters with exactly one thread — 1/8 with #4-40, 1/2 with 3/8-16, 1 inch with 3/4-10 — and the pairing is the standard’s, not a supplier’s choice. There is no 1/2 inch shoulder screw with a 1/2 inch thread and there never was. In the trade the part is also called a stripper bolt, which is the die-set job it was designed for: the stripper plate rides on the shoulder while the thread holds the screw in the punch holder.

The shoulder is two thousandths of an inch under nominal at every size, and that is the clearance strategy. A 1/2 inch shoulder screw has a 0.498 inch shoulder, a 1 inch has 0.998, a 2 inch has 1.998. One rule reproduces the whole column, which is how you know it is a rule. The practical consequence is in the bore: ream to the nominal size and you have handed over two thou of diametral clearance before adding any running allowance, so a locating pivot built that way has play in it that nobody specified. Size the bore from the shoulder’s actual maximum, which this page prints beside the nominal, and add the fit you want on top.

The subtraction that decides whether the joint works is shoulder length minus grip. It has to be positive. The shoulder must reach the tapped face BEFORE the head reaches the rotating part, so that the screw can be taken to full torque while the part stays free. If the stack is thicker than the shoulder, the head lands first and tightening clamps the thing that was supposed to turn — and the joint silently becomes a friction joint with an undesigned clamp load. This page makes that difference the headline for that reason, and it goes negative rather than merely looking odd. The commonest way the error arrives is by counting the tapped plate in the grip. The grip is only what the shoulder passes through: the rotating part, and any washer or spacer under the head.

This is also why you cannot substitute a socket head cap screw, and it is not a matter of degree. A cap screw has no shoulder to bottom against, so there is no length at which torquing it leaves the part free. Either it is loose, in which case the plain part of the shank — which on a cap screw is not ground, not controlled in diameter and partly threaded — wears the bore oval, or it is tight, in which case the part is clamped. People make this swap because the two screws look alike in a bin and because a cap screw is cheaper, and the joint usually works for a while.

The shoulder is the strong member, by roughly two to three times. Its full cross-section runs between 1.92 and 2.76 times the thread’s tensile stress area across the range, so in shear and in bending the shoulder governs and the thread is only retaining the screw. Expect an overloaded shoulder screw to bend or shear at the shoulder rather than strip its thread, which is the opposite of what a cap screw does. One qualification, and it is the reason the table on this page carries an area-ratio column: the margin is much thinner at the small end. At 1/8, 5/32 and 3/16 inch the shoulder is only about a tenth larger in diameter than its own thread, so it buys a fifth to a quarter more area rather than up to double, and at those sizes the hexagon socket — 5/64 or 3/32 inch — is usually the weakest part of the fastener anyway.

And metric shoulder screws exist, whatever you have been told. ASME B18.3 is an inch-series document and has no metric table, which is where the folklore comes from. ASME B18.3.3M, Hexagon Socket Head Shoulder Screws (Metric Series), covers 6.5 to 25 mm nominal shoulder diameters; it dates from 1986, was reaffirmed in 2008, and is now retained under stabilized maintenance rather than being an active standard, which is why it is easy to miss. Distributors stock parts to it, with the same one-size-down rule — a 20 mm shoulder on an M16 thread, a 24 mm on an M20. The dimension table on this page is inch only because that is what three independent sources could be found to agree on; that is a limit of the page, not of the part.

Where to go next on this site. The counterbore and the clearance hole for the head are a subject of their own and the counterbore and countersink page carries the standard depths and the drill sizes. For the running fit on the shoulder itself, the ISO 286 fit calculator will give you an H8/f7 or H7/g6 bore from the shoulder maximum this page prints — which is the right way round, because the shoulder is the fixed member and the bore is the one you are machining. If the tapped depth is the question rather than the shoulder, the thread engagement length page sizes it properly for the thread and the parent material. And if you want to check the shoulder as a beam with a load case other than the mid-grip cantilever used here, the beam stress and deflection page has the seven standard cases with the section properties of a solid round among them.

Frequently asked questions

What thread does a 1/2 inch shoulder screw have?

3/8-16 UNC. The thread on a shoulder screw is always one standard size below the shoulder, and ASME B18.3 fixes the pairing rather than leaving it to the maker: 1/8 takes #4-40, 5/32 takes #6-32, 3/16 takes #8-32, 1/4 takes #10-24, 5/16 takes 1/4-20, 3/8 takes 5/16-18, 1/2 takes 3/8-16, 5/8 takes 1/2-13, 3/4 takes 5/8-11, 1 inch takes 3/4-10, and so on to 2 inch on 1-1/2-6. There is no size at which the thread matches the shoulder. If a listing offers you one, it is not a shoulder screw to B18.3.

Why is the shoulder smaller than its nominal size?

Because it is a bearing surface, and a bearing surface has to be undersize for the bore to be made on size. ASME B18.3 holds the shoulder at the nominal minus 0.002 inch at every one of the fourteen sizes, so a 1/2 inch shoulder screw measures 0.498 inch and a 1 inch measures 0.998. The number matters because it is easy to miss: if you ream the bore to a nominal half inch you have already given away two thousandths of diametral clearance before adding whatever running fit the design asks for, and the pivot has play nobody specified. Size the bore from the shoulder’s actual maximum instead.

How long should the shoulder be?

Longer than the stack it passes through, and by a controlled amount. The shoulder has to bottom on the tapped face before the head touches the rotating part, so that the screw can be taken to full torque while the part stays free to turn. Shoulder length minus grip is the clearance, and it is the headline on this page. Zero or negative means the head clamps the part and the joint is not a pivot. Something in the region of 0.05 to 0.25 mm is a sensible working band for a pivot that also has to locate; a stripper plate that is meant to travel can have far more. Count only what the SHOULDER goes through — the rotating part and any washer or spacer under the head — and not the plate the thread screws into.

Can I use a socket head cap screw instead of a shoulder screw?

Not for a pivot or a stripper, no. A cap screw has nothing to bottom against, so there is no length at which tightening it leaves the part free: it is either loose, in which case the unground and partly threaded shank wears the bore, or it is tight, in which case the part is clamped. The shoulder screw exists precisely to separate those two things — the shoulder sets the axial position and the thread supplies the clamp. The swap happens because the two screws look similar in a bin and because the joint usually survives for a while, which is the worst combination for learning from it.

Is a shoulder screw a stripper bolt?

Yes, the same part under two names. Stripper bolt is the die-set name: the stripper plate rides up and down the ground shoulder while the thread holds the screw in the punch holder, and the plate’s travel is set by the shoulder length against the plate thickness — the same subtraction this page is built around. You will also see shoulder bolt and, loosely, stripper screw. The dimensional standard is ASME B18.3 whatever it is called on the invoice.

Are metric shoulder screws available?

Yes. The standard is ASME B18.3.3M, Hexagon Socket Head Shoulder Screws (Metric Series), which covers 6.5 to 25 mm nominal shoulder diameters. It was published in 1986 and reaffirmed in 2008, and ASME now lists it under stabilized maintenance rather than as an active standard, which is why the usual answer to this question is wrongly no. Distributors stock the parts, and they follow the same rule as the inch series: a 6 mm shoulder on an M5 thread, 20 mm on M16, 24 mm on M20. The table on this page is inch only because the inch figures could be confirmed in three independent sources; the metric ones could not, so they are not printed.

Does the shoulder or the thread carry the load?

The shoulder, by a wide margin, and that is the point of the part. Its full circular cross-section is between 1.92 and 2.76 times the thread’s tensile stress area across the standard’s range, so in shear and in bending the shoulder governs and the thread is only retaining the screw. The practical consequence is which failure to expect: an overloaded shoulder screw bends or shears at the shoulder rather than stripping its thread. Two qualifications. The margin is thinnest at 1/8 to 3/16 inch, where the shoulder is only about a tenth larger in diameter than its thread. And on a pivot the governing case is usually fatigue rather than static strength, because the load reverses every cycle and neither the head fillet nor the thread runout is modelled by a static calculation.

Related calculators

References

ASME B18.3, Socket Cap, Shoulder, Screw and Hex Keys (Inch Series), cited by number and not reproduced. It is the document that makes a shoulder screw a defined part rather than a turned special: it fixes the shoulder diameter, the head, the hexagon socket, the thread and the thread length for fourteen shoulder sizes from 1/8 to 2 inches, and it pairs each shoulder with ONE thread. The pairing is the standard’s, not a choice a supplier makes. What is taken from it on this page is attributed to the catalogues the printed values were read in, below; the arithmetic built on top of them — bearing area, shear area, section modulus, UNC stress area, the grip-length check — is computed here.
Fuller Fasteners, published ANSI/ASME B18.3 table for hex socket head shoulder screws. The source of the fourteen-row shoulder screw table on this page: shoulder diameter, head diameter, head height, hexagon socket, thread, threads per inch and thread length, for 1/8 through 2 inch shoulders. Fuller prints threads per inch alongside the thread, and in this size range the TPI alone identifies the UNC thread, which is how each pairing was confirmed against independent distributor listings rather than taken on trust.
AFT Fasteners, Socket Shoulder Screw Dimensions (Alloy Steel) ASME B18.3-2003 (read 6 October 2026). The independent confirmation of the middle of the shoulder screw table. It reproduces shoulder diameter max and min, head diameter, head height, hexagon socket, thread, TPI and thread length for 1/4 through 3/4 inch, and every one of those figures matches Fuller’s exactly — including the thread length column, which is the one column no other source confirmed. It also gives the mechanical properties: 140,000 psi minimum tensile, 120,000 psi minimum yield.
EngineersEdge, Inch Series Socket Head Shoulder Screws (read 6 October 2026). The second independent confirmation of the shoulder screw table, and the one that reaches the top of the range: it prints shoulder diameter, head diameter, head height, hexagon socket and thread for all eleven sizes from 1/4 to 2 inch, max and min, and all of them match. Its last column is NOT thread length despite how it reads in places: the values are exactly 0.75 × the hexagon socket size at all eleven rows, which is the key engagement and not the thread. Mentioned here because a column read off the wrong heading is how a wrong number gets into a page.
TorqBolt, ASME B18.3 hexagon socket head shoulder screw dimensions page (read 6 October 2026). Read specifically for the three SMALLEST sizes, 1/8, 5/32 and 3/16 inch, which neither AFT nor EngineersEdge carries. It confirms head diameter, head height, hexagon socket, thread and TPI at all three. Two differences are worth recording. It presents the shoulder diameter figures 0.1230, 0.1540 and 0.1850 as the MINIMUM of a window whose top is the nominal, where the other three sources present the same numbers as the maximum; and it gives the 1/8 thread length as 0.157 against Fuller’s 0.156. Its own key-engagement column prints a maximum below its minimum, so that column was not used.
Bolt Depot and MSC Industrial Supply shoulder screw listings, used during the data preparation for this batch to confirm the thread pairings at the three smallest sizes independently of any dimension table: 1/8 inch shoulder to #4-40, 5/32 to #6-32, 3/16 to #8-32. Every threads-per-inch figure matched, which in this size range is enough, because 40, 32 and 24 TPI each identify a single UNC thread.
ASME B18.3.3M, Hexagon Socket Head Shoulder Screws (Metric Series), 1986, reaffirmed 2008. Read from ASME’s own catalogue entry on 6 October 2026, which gives the title, the reaffirmation date, a nominal shoulder diameter range of 6.5 to 25 mm, and the status: no longer an active ASME-approved standard, retained under stabilized maintenance for historical reference. It is cited here for one reason. A metric shoulder screw standard EXISTS, which is the opposite of what is usually said about this part, and a reader who gives up looking because somebody told them the component is inch-only has been misinformed.
McMaster-Carr, shoulder screw listings filtered to the specification ASME B18.3.3M (read 6 October 2026). The evidence that metric shoulder screws are a real catalogue item and not a theoretical one, and the source of the three shoulder-and-thread pairings printed on this page: 6 mm shoulder with M5×0.8, 20 mm with M16×2, 24 mm with M20×2.5. The same one-size-down rule as the inch series, in metric. Note that the 6 mm entry falls below the 6.5 mm bottom of the range ASME’s own catalogue page states for B18.3.3M, which is said on the page rather than smoothed over.
ISO 273, Fasteners — Clearance holes for bolts and screws. Cited by number and NOT used numerically, which is worth being explicit about: its three fit classes (fine, medium, coarse) were not obtainable in a readable published table during this batch, so the clearance holes printed on this page are the close-fit and normal-fit columns from Machinery’s Handbook by way of amesweb, attributed there and not to ISO 273. If you are working to ISO 273 specifically, check its own table: the two systems are close but they are not the same document.

CalcEngines mechanical calculators are for education and design estimates. Results depend on the values you enter and the assumptions stated on the page, and they do not replace a qualified engineer's check. Standards, material grades and manufacturer data change — verify against the current standard and the supplier's own catalogue before you specify, machine or build anything.