Bolt Tensile Strength Chart: Grade Table (Free PDF)

ISO 898-1 property classes list strength of metric bolts. Every metric hex bolt ≥ 5 mm in diameter have this number stamped on its head. Reading from left to right in the table above, look at the marking on the bolt head, which tells you the line you’re going to read across to find the following:

• Proof stress: The stress a bolt must survive with no measurable permanent set. Yield strength is the guaranteed minimum stress at which plastic deformation occur. Tensile strength is the guaranteed minimum stress at which the material break.

The metric stress numbers are in units called megapascals (MPa). Convert a stress number to a force (or weight) by multiplying it by your thread size’s tensile stress area. Find these numbers in the tables below. You’ll need them for calculating proof load and breaking load.

That's what the class notation actualy encodes. The nominal tensile strength is just the first number time 100. So for instance, if it says 8.8, you know that's about 800 MPa. Multiply the first number by the second number, then multiply by 10 to get the nominal yield strength. So 8.8 = ~800 MPa tensile strength. Also, 8*8 *10 = 640 MPa nominal yield strength. 12.9 = 1200 MPa tensile strength, and 1080 MPa nominal yield strength.

This is how I decode the values from the Nominal tensile column. The Min tensile column have a slightly different but more realistic minimum guaranteed by the standard. This is what you use in your calculations.

And beware of class 5.8. This one doesn't have 600 MPa as its nominal tensile strength. It's only got 500 MPa. The minimum is 520. A few quick reference charts does this incorrectly.

The min yield strength column show the stress that will begin deforming the bolt permanently. On low classes (4.6, 5.6), it’s the actual lower yield point ReL. On the cold-worked classes such as 4.8, 5.8, 6.8, it’s the full-size equivalent Rpf. And on quenched and tempered classes (8.8 and up), it’s the 0.2 percent proof stress Rp0.2.

The proof stress Sp column show even lower values. It’s the stress level your full bolt can withstand without any measurable amount of permanent set during a proof test. You’ll use these numbers to compute published proof loads. You can use those proof loads to find your preload and tightening torque recommendations for most purposes.

The table has two rows under Class 8.8. ISO 898-1 specifies different minimum values for sizes above and below M16. For sizes over M16, the minimum tensile strength is 830 MPa with a yield of 660 MPa. Up to and including M16, it's 800 MPa with 640 MPa yield.

The Vickers hardness column give the acceptance range for the finished bolt. This is the fastest thing to check in the workshop if the mark cannot be read. The Elongation column indicate minimum ductility from machined test pieces. Class 12.9 has 8 percent elongation. Class 8.8 has 12 percent elongation. So 12.9 is much more unforgiving of any prying or bending loads compared to 8.8.

There’s often confusion when people mistake an imperial-grade bolt for a metric class. An 8.8 on a metric bolt refers to 800 MPa (about 116 ksi). That’s almost Grade 5 SAE, not Grade 8! Similarly, 10.9 in the metric system would be about 1040 MPa (151 ksi). That’s roughly Grade 8 SAE.

Radial line markings belong to the imperial system in the next table. Numbers such as 8.8 or 10.9 is never imperial; they’re always metric. Small-diameter bolts under 5 mm might not even get marked at all. And many lower classes doesn't get marked either. You should of then assume the lowest possible class unless it indicates otherwise on the package.

Stainless steel bolts are not included in this table. They have markings such as A2-70 and A4-80. This is ISO 3506. Neither are set screws. Those are classified under ISO 898-5, with class numbers such as 14H and 45H.

Structural bolting assemblies to EN 14399 is not included. Nor are nuts. Nuts have single number classes such as 8, 10 and 12 (ISO 898-2).

Note that the table applies only to room temperature. Above about 300 degrees C, strength falls off sharp. Standard classes are not rated for creep service at those temperatures.

This table provide a quick reference for the strength of inch-series bolts rated to common ASTM specifications as well as to SAE J429 specs. It covers bolt hardware grades from 1 through 8.2.

Look up your grade in the head marking or the spec you used to buy the bolt. Next look at whether your diameter is included in the size range column. Finally note the proof, yield and tensile stresses in ksi.

Ksi are thousands of pounds per square inch. A ksi = 1000 psi. The approx MPa column converts the minimum tensile strength so you can compare it against the table above in metric.

The quickest indicator on a bolt is its head markings. If there are no radial lines, it’s either Grade 2, Grade 1, or some other grade not specified here. If there are three radial lines, it’s a Grade 5. If there are six radial lines, it’s a Grade 8. The easy answer? Lines + 2 = grade.

All graded bolts require a mark from the manufacturer as well. In place of radial lines, structural bolts is stamped by spec. They are stamped A325 or A490 in either raised or depressed lettering. Grade 307 are A307 bolts and are stamped 307A or 307B. An unmarked head is legally only a 60 ksi class fastener at best: never assume more. Never assume stronger.

Within each grade look out for the size step. For instance, in Grade 2 it’s 74 ksi tensile strength up to 3/4 inch. At 7/8 inch the strength reduce to 60 ksi. That’s what accounts for the lower proof load of a 7/8 inch Grade 2 bolt compared to a 3/4 inch one in the UNC table below.

Likewise, Grade 5 go from 120 ksi to 105 ksi at sizes larger than an inch. Grade 8 stays steady at 150 ksi throughout.

In the old A325 spec, they had 105 ksi over 1 inch. Those bolts are all still in service and are kept here as their own row since there is no longer a separate specification. It was consolidated into ASTM F3125 in 2016. New Grade A325 bolts are 120 ksi now at all diameters from 1/2 to 1 1/2.

Proof Stress Column, This represents the stress a full-size bolt must hold with no permanent set. It’s this number that gets used as the basis for published clamp load and proof loads tables.

Typical assembly preload range from 75 to 90 percent of proof.

Where there are blank cells here, the standard does not define a proof test for those grades. Instead, it only defines a minimum tensile number such as what is done for Grade A307. That means there are no torque tables based off proof load for A307. Don’t make one up.

Grades 5.1, 5.2 and 8.2 is equal to their parent grades in strength. But they have a more tightly defined size range. And for boron steels, they have a lower temperature tolerance limit; around 250 degrees C max in service.

Never confuse specs, and do not mix systems. Metric 8.8 is approximately equal to SAE Grade 5. Metric 10.9 is about equal to Grade 8.

Never hot-dip galvanize a Grade 8 or A490 bolt because of the risk of hydrogen embrittlement. Always use Grade 5 or A325 type if galvanized is needed.

There’s also a rarely seen SAE J429 Grade 7 that is heat treated, then roll threaded to yield 133 ksi tensile strength. Few suppliers carried this grade. We did not include values for it as we couldn’t cross-check them.

This does not cover socket head cap screws or stainless bolts. Also excluded: Metric ASTM F568M classes, anchor bolts (F1554) beyond the strength summary, or socket head cap screws.

The table translates those stress classes into real-world forces. For each of the coarse metric threads ranging from M3 through M39, it shows the minimum ultimate tensile load and the tensile stress area. These are expressed as a load in kilonewtons based off property class.

Locate the size of thread you’re using in the left-hand column. Look across the row until you see the kN value below your bolt’s class. That’s the lowest force at which a bolt can fail during an otherwise pure tension test while still being considered standard compliant. A kilonewton is approximately 225 lbf or 102 kgf.

The Pitch column confirms you are looking at the regular coarse thread. An M12 coarse bolt is a 1.75 mm pitch. Fine pitch threads has slightly larger stress areas and higher load values, so dont use this table for an M12x1.25 bolt.

Look at the As column, which gives the effective cross-sectional area of the threads in square millimetres. That’s based on the average diameter between the minor diameter and the pitch diameter. On purpose, that’s less than the cross-sectional area of the shank. The specific value depend on the bolt size.

All of those numbers in the table are nothing more than As times the lowest tensile strength for the class. They’re rounded off to three significant figures. You could go further with this and apply it to any class or size that isn’t covered in the table.

The above are breaking loads. Not working loads. A bolt at its tabulated value is going to fail. But it will yield and stretch permanent long before then.

To select an assembly preload you look in the following table of proof loads. Then you apply the safety factor of your code for the working load of your design. In machine design, common practice is to keep the service load less than about half the proof load. Other codes (structural) has their own partial factors.

Also the values are based on a full bolt loaded at room temperature under slow axial loading. Real capacity gets reduced significantly by shear loading. Capacity is also reduced by bending due to non-flat seating. Elevated temperature and fatigue reduce capacity substantial too.

See those blank spaces? Here is where it notes exceptions in footnotes.

Class 9.8 applies through M16 only. There is no value beyond that size.

For class 8.8 in structural bolt applications, ISO 898-1 gives slightly lower load numbers: M12 (70.0 kN), M14 (95.5 kN) and M16 (130 kN). That corresponds to structural nut standards. The standard values are the ones in the main cells.

At M8 and M10, some classes have been given reduced values as per ISO 10684 for hot-dip galvanized bolts with undersized 6az threads. Note that ISO 898-1 terminates at M39.

Commercially available larger sizes do exist: M42 to M64. Their guaranteed loads must be obtained either from the maker or by reference to a project spec, not from this table.

This does not apply to threaded rod of unknown class. This does not apply to stainless steel. The strength of stainless steel varies, but the stress areas match the ISO 3506 standard.

This does not apply to bolts with corroded or damaged threads. It reduces the effective stress area.

The class system also applies down to M1.6 if that’s the size you require. Below M3 there is no load table published in ISO.

Instead, multiply the stress area by the class strength.

The second table, above, show the proof load in kilonewtons for coarse (M3 to M39) threads. Use it along with the previous breaking load table.

Proof load is the force that a completed bolt should of hold for 15 seconds. The completed bolt must let go of the proof load without any permanent elongation. By definition, this represents roughly 88 to 94 percent of the minimum yield strength.

In other words, it’s the best number when designing a bolted joint. It’s what you make your assembly preload and tightening torque recommendations out of. For reused bolts, the preload is normaly 75 percent of the proof load. For permanent joints, it's set as a fraction up to 90 percent of this value.

The columns are arranged the same way as the breaking load table. The Pitch column verifies that it's a coarse thread. The Stress area As column show the mm² size of the thread cross section. Then each class column are simply As times the proof stress Sp from the above table.

For classes 4.6 through 12.9 this stress start at 225 MPa and increases to 970 MPa for class 12.9. The numbers were rounded off to only 3 significant figures.

Proof stresses also vary depending on the class. So the proof load is not going to be a known percentage of the breaking load. Class 4.6 will have a proof load around 56 percent. Class 8.8 is around 73 percent. And class 12.9 around 80 percent.

A common shortcut is T = K x F x d. In this formula, K is the nut factor, F is the preload you want to achieve (such as 0.75 times Fp), and d is the nominal diameter in metres.

For plain dry steel, K is about 0.20; for lightly oiled surfaces, it’s 0.15 to 0.18; and for anti-seize or moly lubricants, it’s down to 0.10 to 0.12. In other words, the variation in K pretty much render any one torque figure worthless unless accompanied by some description of the lubrication condition. So that’s why they publish the loads instead of torques on this chart.

The same caveats apply as with the breaking load table. This only applies to coarse thread. Values for 8.8 are reduced where used structurally (footnote). Some sizes has reduced values for galvanized 6az threads. Class 9.8 stops at M16, and the table ends at M39. It ends at M39, as per ISO 898-1.

Proof load has one additional trap that is unique to it. Protection is only against permanent stretch when tightened. If there are substantial working loads over and above preload then a full joint calculation are required. Further analysis is required if vibration is an issue or temperature is elevated. Being under proof at assembly doesn't make the joint safe.

Any proof load you run across on paper that doesn't match up in the third significant figure (slightly) in this table, check where it came from. Was it calculated using the nominal proof stress like ISO? Maybe it's a rounded percentage off yield?

Differents in the third significant digit should be expected between sources. These ISO numbers I have provided above are the contractual numbers for a certified bolt.

Above is a table of actual forces on inch coarse (UNC) thread bolts ranging in size from 1/4-20 to 1 1/2-6. These are common SAE J429 grades of bolts.

Column one indicate the number of threads per inch and the nominal size. Next is the stress area, which shows the effective cross-section at the threads in square inches. After that there is two columns per each grade.

Proof load indicate the max force without any permanent set. That is the basis for our torque and preload values. Minimum Tensile load is the amount of force below which the bolt should of not fail. Lbs-force are all units.

Multiply the grade stress by the stress area to find each value. Values are rounded and published by SAE. So 1/2-13 Grade 5 at 85,000 psi x 0.1419 in2 = 12,100 lbf Proof Load.

You can recreate a value for a spec or grade not listed. Grade 5 compares to A449. Grade 8 roughly compares to A354 BD. Multiply its stress by the stress area column to make a value.

Here you see that the grade size steps make sense when looked at on a table. Look at the Grade 2 proof column, go down it. From 3/4 in to 7/8 in there is a drop from 18,400 lbf to 15,200 lbf. Why? Because Grade 2 falls from 74 to 60 ksi above 3/4 in.

So, the larger Grade 2 bolt can actualy be weaker in proof than the size below it. The same thing occurs with Grade 5 at 1 in. Grade 5 drop from 120 to 105 ksi above 1 in.

Grades 1 & 8 maintains the same stress across the whole size range. That’s why Grade 1 and Grade 2 has equal columns from the 7/8 mark.

Assemble by proof load; tensile load is for ultimate capacity checks.

Torque is estimated as T = K x F x d.

The same nut factors apply as they do with the metric table.

Set it at 0.20 for dry bolts or 0.15 for oiled bolts.

The Tensile Load columns is for checking ultimate capacity and failure analysis. They are not for setting torque.

Preload is typically 75 to 90 percent of proof. Other clamp load tables is typically exactly 75 percent of those proofs.

Sizes are less than 1/4 in. aren't covered by this table. Full-size (J429) starts with 1/4 size. Only Grade 5.1 SEMS values covers small machine screw.

Sizes are greater than 1 1/2 in. aren't covered by this table. Above that, use ASTM specs like A449 up to 3 in or A354 up to 4 in.

Next table covers fine threads. No structural A325/A490 bolts here. These bolts are produced only from 1/2 in. and up.

Turn-of-nut or tension-control methods is used to tighten them. Structural A325/A490 bolts are tensioned by turn-of-nut or tension-control methods different than using these generic proof loads.

Above is the table of Minimum Tensile Load and Proof Load in pounds-force for Fine Threads (UNF), grades 1, 2, 5, and 8. Notice that each column is listed in the exact same order as the UNC table. The first two columns list out the size of thread and number of threads per inch. Next is the tensile stress area in square inches. Then you’ll find a pair of proof load and minimum tensile load for Grade 1, 2, 5, and 8 thread.

First, for a given size of diameter, a fine thread is shallower; therefore it has greater stress area. Thus any load on that bolt will be about 9 to 14 percent greater than on an UNC bolt of equal nominal size. A 1/2-13 bolts proves at 17,000 lbf while a Grade 8 1/2-20 proof is at 19,200 lbf.

The finer threads is less prone to loosen up and allow finer adjustments with preload. They do have some disadvantages however: they are harder to get started (cross threading), more susceptible to damaged threads, and take longer to install.

In the aerospace industry and automobile manufacturing, fine threads are used as a matter of course, but in general construction they are uncommon.

Never combine series of threads in your substitution. A 1/2-13 bolt will begin with a 1/2-20 nut, but it won't go all the way on, stripping out instead. Both the TPI (threads per inch) and the diameter must match the values shown in the table's row.

If you have 1-inch thread, then it's either 1-14 (UNS special), or it's 12 TPI for proper UNF. That's why that row is in the table. It is J429 tabulated. It is also very popular on older axle hardware and spindles on machinery.

The size steps for grades are the same as on the coarse table, so it’s got all those counter-intuitve dips. The Grade 2 load drops to Grade 1 levels with a load at 7/8 inch. From there up, both columns is equal. Then the next drop is on Grade 5, which is above 1 inch.

As usual, set your preload from the proof column, and use 75 to 90 percent, which is typical. Stay inside your design code’s factor of the tensile column on working loads. These figures assume that the threads are at room temperature, clean, and not damaged, and that the thread is used only in pure tension.

The UNEF and 8UN constant-pitch series is excluded from this table.

Metric fine pitches are also excluded. Use the metric table with fine pitch stress area.

Left-hand versions are not included because thread hand does not affect strength values. Stress areas vary by series, so verify the series in question.

For sizes larger than 1 1/2 inch or grades other than J429, refer to the stress area formula and governing ASTM spec. You should of checked the stress area calculation if the size is different than the table.