In designing mechanical parts in an assembly, sometimes the mating surfaces are raised faces. I've been told it's to provide a better seal. However, for piping (flanges), it seems to be the consensus that a raised face flange does not provide any extra seal functionality, but has more to do with the manufacturing method (forged vs cast).

My question doesn't necessarily apply just to piping flanges, but to designing parts in general. Is it desirable to use a raised face on mating parts, and when (always, sometimes, etc)?


In general you would raise mating surfaces in order to make them easier to machine.

Having the surface raised above the surrounding material means that you don't have to worry so much about tools colliding with edges of the feature and it's fairly intuitive that skimming a raised surface is a simpler operation than machining a recessed pocket and it minimises the amount of material that you need to remove.

Similarly if you are machining a raised surface with a bit of 'meat' to it you can concentrate on surface finish and flatness without too much concern about the precise depth of cut and you have some material to work with if you need to rework a part during original manufacture or in-service maintenance/repair.

A raised surface also means that the overall dimensions of mating parts can be less precise as slightly oversized parts will just overlap rather than interfering with anything.


In many cases, it's done for the sake of precision.

As an example, an engine block is often sand cast. All mating surfaces in the casting master are made oversize, with the intention being that they will later be milled down to their final size. Since milling is orders of magnitude more precise than casting, this yields a more consistent result — where required — while maintaining the lower overall cost of the casting process.


Surfaces on a two part mated assembly that is required to provide sealing must be designed so that the seal acts as a seal. While that sounds obvious, it seems to have escaped too many designers and designs which attempt to rely on properties of the jointing materials are too common. While such designs may work when new, "properties of the jointing material" is often time variant, and Murphy ensures that in most cases the variation is in an undesirable direction. (One exception, in many cases, are the "silicone rubbers". If these seal initially and use materials properly matched to the jointed materials, lifetimes of 20+ years are not uncommon.)

In the case of sealing against a pressure difference a simplistic perspective is that the force per area at the seal surfaces from the material in the pipe must be lower than the force per area exerted by all sources of sealing pressure. That sentence may sound strange but is written that way because there may be more than one source of sealing pressure.

The obvious source of sealing pressure is the bolts retaining the join.
An obvious way of ensuring that bolting "psi" (lbf/in^2, Pa, ...) is maximised is by minimising the area so there is "more pressure to go around" for a given bolting force.

Raised flanges allow formal sealing areas to be designed and area to be minimised.
Without such raised areas sealing psi is reduced.


Another less obvious and exceedingly useful means of exerting sealing pressure is to use the internal pressure to create its own sealing forces. This is the principle used by "O rings" which use pressure differential to deform a flexible ring in such a way that the sealing psi exceeds the leakage psi from the same pressure source. While O rings can be easily designed by following standard rules and tables, they in fact involve very black magic indeed, and numerous low cost designs that appear to employ O rings in fact just use squashed rings of flexible material. The difference between a true O ring seal and a bit of squashed rubber is proper design and manufacture and a relatively small extra $ outlay. The difference in performance is incomparable.

  • $\begingroup$ Surely you mean simply "pressure" and not "pressure per area?" $\endgroup$ – Air Mar 16 '15 at 18:22
  • 2
    $\begingroup$ @Air Thanks - sort of - I've changed the term to"force" as what I meant was that the available force per area from clamping etc exceeds the force per area from gases. Arguable "force per area" = pressure so that term could be used BUT gas psi is developed elsewheere and if you halve area you get half the force so same psi (of course) BUT eg bolting force is ~= constant as you reduce mating areas so halving flange seating area doubles force per area so doubles sealing pressure. Just saying Pseal > Pinternal hides this crucial relationship. $\endgroup$ – Russell McMahon Mar 17 '15 at 3:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.