Busbar hole patterns and what DIN 43673-1 actually specifies

Search for DIN 43673 and every result on the first page wants to sell you the document. Nobody explains it. The standard is four pages of German from February 1982, and the number turns up on customer drawings constantly, usually as a bare reference with no dimensions attached.

What the standard covers

DIN 43673-1

, Stromschienen-Bohrungen und -Verschraubungen; Stromschienen mit Rechteck-Querschnitt, published in English translation as "Drilled holes and screw connections for busbars; rectangular cross-section busbars". Four pages. It specifies the location and size of the drilled holes and gives base data for the design and construction of the screw connections. It applies to direct current and to alternating current up to 60 Hz.

Its scope is tied to two other DIN documents: it applies to the rectangular bars whose continuous current ratings are given in DIN 43670 (aluminium, 1975-12, ten pages) and DIN 43671 (copper, 1975-12, nine pages). The family splits cleanly. DIN 43671 tells you how much current a given bar section will carry; DIN 43673-1 tells you where to put the holes in it. Both are still listed as current by DIN Media, which is unusual for documents of that vintage.

Two things follow from the scope wording. The standard is about the connection, not only the drilling, and the "base data for the design of the screw connections" is the half people forget. And the 60 Hz ceiling is real: above it, skin and proximity effects in the joint zone redistribute current around the holes enough that the pattern stops being the right one.

DIN 46433 is not a busbar hole standard

This correction gets its own heading because the miscitation is everywhere, including in tender documents and in supplier drawings that get copied forward for years.

DIN 46433

is titled Flachdrähte und Flachstangen, gezogen, mit gerundeten Kanten — Maße: "Rectangular wires and rectangular bars, drawn, with radiused edges; dimensions". Eight pages. It is a dimensional standard for drawn rectangular wire and bar: width, thickness, edge radius, tolerances. It says nothing about hole positions, hole diameters or bolted connections. It has also been withdrawn, and is superseded by DIN EN 13601.

If a drawing you receive says "holes to DIN 46433", the drawing is wrong. The citations that person meant are DIN 43673-1 for the hole pattern and DIN 43671 for the current rating, and if they were reaching for the material spec, that is EN 13601, the same document that replaced DIN 46433. Raise it before you cut metal, because a drawing that miscites one standard usually miscites the temper as well.

Where the hole pattern meets assembly verification

Hole patterns look like a fabrication detail. They are not, because the joint they produce sits inside the scope of the assembly standard the finished panel is verified against.

IEC 61439-1 is explicit. In the temperature-rise test, clause 10.10.2.3.3 requires that "particular attention shall be given to joints in conductors and terminals within the main circuits". Joints are named measurement points. Clause 10.10.2.3.7 a) requires that where main busbars are tested separately the tested length is at least 2 m and includes one joint where the bars are extendable, and the short-circuit clause 10.11.5.3.3 fixes that length at (2 ± 0,4) m on the same condition.

Then the acceptance criterion. After the short-circuit test, clause 10.11.5.5 states that "there shall be no loosening of parts used for the connection of conductors". A joint that survives thermally but backs off under electromagnetic force fails the assembly, and edge distance and bolt sizing are what decide that.

Edge distance

The standard fixes hole centres relative to the bar edge. Two independent things are being balanced.

Mechanically, the ligament between hole and bar edge carries bearing load from the bolt during a fault. Peak withstand current drives adjacent bars apart or together at forces that end up reacted through the bolt group. Too little edge material and the ligament yields, the hole elongates, preload drops and the joint loosens. Outright tear-out is rare in copper of sensible thickness; the real failure is progressive elongation followed by loss of contact pressure, which is worse because it stays invisible until the thermal survey picks it up.

Electrically, the ligament is a neck. Current has to detour around the hole, so the material between hole and edge carries a higher density than the bulk of the bar. Put the hole too close to the edge and you have built a local hot spot into the most highly stressed part of the joint.

Manufacturing adds a third constraint. When you punch rather than drill, the die needs material around it to react against. Punch with too little edge distance in hard-temper copper and the edge bulges outward, the ligament work-hardens, and the bar no longer sits flat against its mate. As shop practice rather than a normative figure, keep the clear distance from hole edge to bar edge at least equal to the bar thickness, and 1.5 times it in hard temper. Our punching process page covers the tooling side of the same constraint.

Pitch, and current crowding around the holes

The Copper Development Association's Copper for Busbars (Publication 22) gives the clearest published treatment of what holes do to a joint. Joint resistance has two parts: the streamline or spreading resistance caused by current having to divert through the overlap, and the contact resistance at the interface itself. Bolt holes make the first one worse, and the estimate CDA gives depends on the hole diameter and on n, the number of holes measured across the width of the bar.

That single variable settles a question fabricators argue about. Holes should be placed in line along the length of the joint. Staggering them across the width increases resistance, because each additional hole across the width is another obstruction in the same current path. Two holes in line along a 60 mm bar are a better joint than two holes offset diagonally, even though the second arrangement looks like it spreads the clamping load more evenly.

The overlap length is governed by the same effect. The streamline penalty falls very steeply as the overlap-to-thickness ratio rises to about 2, then improves only slowly out to a ratio of 10. Past that there is nothing left to gain. CDA's conclusion: the overlap only needs to be long enough to accommodate enough bolts to reach the required contact pressure. Long overlaps waste copper and add weight without buying conductivity.

Two refinements from the same source are cheap to implement and rarely used. Cutting a longitudinal slot through both bars in the overlap reduces contact resistance by 30 to 40 %, because it evens out the contact pressure in each leg of the joint. Angling the bar ends at less than 45° reduces initial joint resistance by about 15 % and slows the rate at which joint resistance climbs under load cycling by a factor of 1.3 to 1.5.

Hole diameter

Hole diameter trades clearance for the bolt against tolerance stack-up and against the contact area you are removing.

Clearance is more generous than people expect. CDA's long-standing guidance pairs an M10 bolt with an 11.5 mm hole and an M12 with 14 mm, rising to 15 mm on wider bars. That is 1.5 to 3 mm of slop, and it is deliberate: it absorbs accumulated pitch error across a five- or six-bolt pattern in two bars that were produced on different days, possibly on different machines. Tighten the clearance and you buy yourself joints that will not assemble.

The cost is contact area. Every hole removes copper from the overlap, and specifically from the region where clamping pressure is highest. A 14 mm hole in a 100 mm bar takes 14 % of the width out of the section at that point. That is why the answer to a marginal joint is more bolts of moderate size rather than fewer large ones. Bolt count buys pressure uniformity; bolt diameter buys section loss.

Hole diameter also interacts with the tool. Punching a hole smaller than the material thickness is hard on tooling in copper and hole quality suffers. It rarely bites at M10 and above in 5 to 12 mm bar, but it does on thin control bar with small fixing holes. The punching force calculator will tell you what a given diameter costs in tonnage before you commit the pattern.

Bolt size, torque and the washer stack

CDA's Table 22 has been the working reference for decades. The pattern runs roughly like this: 25 to 30 mm bar takes two M8 at 17 N·m; 40 mm takes two M10 at 28 N·m; 50 to 80 mm moves to M12 at 45 N·m, two bolts at 50 mm and four at 80 mm; 100 and 120 mm bar takes five M12; 160 and 200 mm bar takes six to eight M16 at 91 N·m. Sizes in service run M6 to M20, four or six bolts being the common case.

Torque is a proxy, and a poor one. Contact pressure cannot be measured in the field, so it is inferred through T = K·F·D, where K is the nut factor. CDA gives K as 0.20 to 0.22 dry, 0.19 to 0.21 with a contact-aid compound, and 0.15 to 0.16 with a molybdenum-disulphide boundary lubricant. Read that as: the same wrench setting on a lubricated thread produces roughly 30 % more preload than on a dry one. If your torque figure came from a table and your assembly instruction says to grease the threads, the two do not belong together.

What you are aiming at is contact pressure across the overlap. CDA's numbers: below about 7 N/mm² is not advisable, above 10 N/mm² is preferred, and above roughly 30 N/mm² there is little further gain. Only around 1 % of the apparent overlap area actually conducts, the load passing through discrete asperity contacts, which is why surface condition matters as much as bolt count and why bars should be flat and freshly abraded at assembly.

The washer stack does real work. Large thick washers spread the bolt load away from the hole and widen the region of useful pressure; CDA pairs M10 with a 24 mm washer 2.2 mm thick and M12 with a 28 mm washer 2.7 mm thick. Where high-tensile steel bolts are used, differential expansion becomes the problem. Steel expands at about 11.1 × 10⁻⁶/°C against copper's 16.5 × 10⁻⁶/°C, so bolt tension climbs as the joint heats and can be driven past the 95 % of proof stress CDA sets as the ceiling. Disc-spring (Belleville) washers absorb that increase; aluminium bronze CW307G bolts at 16.2 × 10⁻⁶/°C avoid it by nearly matching copper.

One counter-intuitive point. CDA does not recommend plating copper-to-copper joint faces unless the environment demands it, because soft plating can flow at elevated temperature and relax the contact pressure you worked to establish.

When the drawing says "to DIN 43673" and nothing else

This is the common case, and it is a commercial problem before it is a technical one.

Ask which bar the pattern belongs to. The standard indexes hole location and size to the rectangular bar sections rated in DIN 43670 and DIN 43671. If the specified width and thickness is not one of those sections, the reference does not resolve and the customer has to give you dimensions.

Ask whether they mean the pattern or the joint. Many drawings citing DIN 43673-1 want only the standard bolt pattern, for interchangeability with existing switchgear. Some want the screw-connection data too, which pulls bolt grade, washer arrangement and torque into your scope of supply. Those are different quotations.

Get one physical reference. A sample bar, a marked-up sketch or a photograph of the mating equipment settles more ambiguity in five minutes than a week of correspondence. The interface holes on the customer's own switchgear are the real specification, whatever the drawing says.

Put your interpretation on the drawing you return, and get it signed. State the hole diameter, pitch and edge distance you intend to produce. If the customer holds the standard and you do not, they can check it against the source in a minute.

Buy the four pages if the pattern recurs. The document costs less than a scrapped batch of 200 mm bar. Order it from DIN Media or your national standards body rather than a reseller aggregator.

Getting the pattern onto the machine

Once resolved, the pattern becomes a tooling question. A recurring DIN pattern across several bar widths means several hole diameters, and the constraint is how many live in the turret at once. A punch and shear line such as the EMAC-BP-60 handles a fixed tool set well. Where the mix is wide, a processing centre like the IMAC-CENTER 60 with 24 tool stations, extendable to 36, takes the changeover out of the equation and holds ±0.05 mm positioning, which is what keeps pitch stack-up inside the clearance you allowed.

The rating side of the same job belongs to DIN 43671, whose tables are referenced at 35 °C ambient and 65 °C bar temperature with correction factors for other conditions. A joint that is thermally marginal because the bar was rated without those corrections will not be rescued by a good hole pattern. Our ampacity calculator applies them rather than hiding them, and the bolted joints article goes further into surface preparation than there is room for here.