A reducing flange carries the flange ring of one pipe size and the bore of a smaller one. Put another way: it bolts to a larger flange and welds to a smaller pipe. That one sentence answers most of the questions people have about them.
What comes from where
Under ASME B16.5 the split is clean. From the larger nominal size: outside diameter, thickness, bolt circle, number and size of bolt holes, and the facing. From the smaller size: the bore, and on a weld neck the hub diameter and length through hub.
Everything that has to mate with the adjacent flange comes from the larger size. Everything that has to mate with the pipe comes from the smaller size. There is no third set of dimensions and nothing is invented.
What it is not
A reducing flange does not reduce the pipe. This is the most common misunderstanding and it leads directly to the second most common one.
The pipe welded to a reducing flange stays at the smaller size. The pipe on the other side of the joint stays at the larger size. The flange simply lets the smaller line terminate in the larger bolt pattern. What has been eliminated is a separate reducer fitting and one weld.
It is also not a reducer. A reducer changes pipe size gradually over a tapered length. A reducing flange changes it instantly, at the flange face. Mechanically those are opposite behaviours, and the difference matters a great deal as soon as the flow has any energy in it.
The thickness detail
People expect a reducing flange to be lighter than a full-size flange, because the bore is smaller. It is not. It keeps the full thickness of the larger size.
The reason is load path. Bolt load acts at the larger bolt circle, and the moment arm from the bolt circle to the gasket is a larger-size dimension. The flange has to carry a larger-size moment, so it needs larger-size thickness. The smaller bore changes nothing about that.
So a reducing flange is heavier than the small-size flange it replaces, not lighter. On big sizes that weight is occasionally the argument that sends a designer back to a reducer plus a standard flange.
Which types exist
- Reducing slip-on. The most common. A bore through a thick ring, so the geometry stays sensible. See reducing slip-on flanges.
- Reducing weld neck. The demanding one, because the hub has to match the smaller pipe while the ring stays large. See reducing weld neck flanges.
- Reducing threaded. Tapped for the smaller pipe. Small bore and retrofit work.
- Reducing socket weld. Socket bored for the smaller pipe, fillet welded.
- Reducing lap joint. A loose larger-pattern backing ring behind a smaller stub end.
- No reducing blind. A blind has no bore. A blanked larger pattern is just a larger blind flange.
Why anyone buys one
- An existing bolt pattern. The mating flange is installed and nobody is redrilling it.
- Space. A reducer plus a flange needs axial length that is not always available.
- One fewer weld. On service where every joint is radiographed and documented, deleting a weld is a genuine saving and a genuine risk reduction.
- Equipment mismatch. A pump, valve or vessel nozzle that is larger than the line serving it.
Why people regret one
Almost always the flow. The abrupt step generates turbulence, and turbulence costs pressure, erodes metal, and on pump suction triggers cavitation. It also ruins the velocity profile that a flow meter depends on.
The other regret is over-reduction. Two nominal sizes is the practical working limit. Beyond that the weld neck hub runs out of material and the flow step becomes severe. See how many sizes can a reducing flange drop.
Ordering one correctly
Two sizes on one line item is where orders go wrong. State both and label them.
- Flange size first, then bore size, and say which is which
- Pressure class, which belongs to the larger size
- Type and facing
- Schedule or wall thickness of the smaller pipe. The most commonly omitted item, and on a weld neck it is essential
- Material grade and specification
- Design conditions, so we can flag a bad application
More on that in specifying a reducing flange on an RFQ.