While it is true that the MIL-B-5087 class H bond value of 100 mOhm is somewhat arbitrary, inability to meet this requirement should be viewed with the greatest concern. Consider the technical reasons for imposing a class H (shock hazard) bond requirement. The ultimate rationale is to avoid a personnel shock hazard. This is accomplished by having a low enough bond resistance (the 100 mOhm value) that a short to case will clear a fault. In the case of 28 Vdc power, a mil-spec bond would draw 280 Amps. That should suffice to trip any conceivable single equipment breaker. With 115 Vac 400 cycle power, a mil-spec bond would draw 1150 Amps rms, and no more need be said. In actual point of fact, the above amperages were calculated neglecting source resistance and that is wrong. Source resistance is important to shock hazard calculations in other ways. If a breaker were to stick (not open under fault conditions) then the mil-spec bond has two further duties to perform. The bond must hold the case with the short to a potential at or below that deemed a safety hazard. For ac power, that is typically 10 Vrms in a dry area, but as low as 1.25 Vrms in a wet area. Clearly the source resistance must be known to design a bond resistance that will just meet such criteria. But that much detail in the design is not necessary, if we also realize that there is one further constraint on the bond path. In the event of a stuck breaker, the bond path must have more ampacity (current carrying capability) than the feeder path. Since the bond path is to structure, and the feed path is a long wire, it is always easy to build a bond that has a low resistance relative to the feed path. If we also assure that the bond path is implemented with a wire gauge size numerically smaller than the feeder (e.g., feeder is AWG 20, bond path is AWG 18) then it is guaranteed that the under stuck breaker conditions that the feed path will open (fuse) before the bond path does. If the bond path is one tenth the length of the feed path and the wire gauge has higher ampacity, the bond path is guaranteed to have less than 10% of the feeder resistance, which will provide the safe equipment case potential rise during the time the fault is not cleared (wet areas would require a 1% criteria, but it should be clear this is still easy to meet). In the above, I have assumed that structure resistance is small with respect to the bond or feed path resistance. For conventional aluminum aircraft - the cited bond value comes from MIL-B-5087 which is an aerospace requirement - that is true. An Al aircraft has a milliohm or sub-milliohm resistance from tip to tail.
I have never imposed a class H type bond as an arbitrary value. The above considerations, suitably implemented, fully address the problem and provide for a practical solution. The one thing I always check and never allow is the use of a low ampacity PCB trace as part of the fault current path. ---------- From: "k3row" <k3...@eurobell.co.uk> To: <emc-p...@majordomo.ieee.org> Subject: Earthing and Bonding List-Post: emc-pstc@listserv.ieee.org Date: Tue, Jul 2, 2002, 4:04 PM I have been looking at an older (but currently in use) item of military equipment and reviewing its safety. It is apparent that the bonding resistance, for one reason or another, is greater than the classic 0.1 ohms. In attempting to derive a view as to the safety of this item of equipment I am aware that I am not very clear as to the basic reasoning and assumptions behind the origin of the 0.1 ohms figure. So can anyone help me out and explain the reasoning behing the value of 0.1 ohms ? Dave Palmer Thales Sensors (UK)