Gaskets ruin the arithmetic
Why a compressible layer invalidates the standard calculation.
A gasket ruins the standard bolted-joint arithmetic because that arithmetic depends on the clamped material being far stiffer than the bolt, and a gasket is deliberately the opposite of stiff, a genuinely soft, compressible layer that changes both how much of an external load actually reaches the bolt and how much clamping force the joint quietly loses over time even with no external load at all.
Why a soft layer changes who carries the load
The favourable load-sharing behaviour covered earlier in this set relies on the clamped material being considerably stiffer than the bolt squeezing it, so that the stiff material absorbs almost all of any shared movement and the bolt's own tension barely changes. A gasket is deliberately engineered to be soft and compressible, precisely so it can conform to small surface irregularities and seal effectively, and that same softness means it can no longer play the stiff partner's role in the joint's load-sharing arithmetic, leaving a meaningfully larger share of any external load to actually reach the bolt directly rather than being absorbed elsewhere.
This is a genuine tension at the heart of gasket design, since the very softness that lets a gasket seal well against an imperfect surface is the identical property that undermines the favourable stiffness ratio the rest of a bolted joint's fatigue resistance depends on, which means a gasketed joint can never fully recover the fatigue advantage an all-metal joint enjoys, however well every other part of the joint is designed around it. Some gasket designs try to resolve that tension directly, embedding a solid metal core or a set of rigid stops within an otherwise soft sealing material, giving the joint something stiff to bottom out against once compressed to its working thickness and recovering part of the favourable load-sharing arithmetic without giving up the soft material's own sealing conformity where it actually touches the mating surfaces.
The foam-cushion comparison
Squeezing a firm wooden block between two rigid plates with a clamp barely deflects the wood at all, but placing a soft foam cushion between those same two plates instead lets the clamp's screw be turned a great deal further before reaching the same clamping force, since the foam does almost all of the compressing that the wood would otherwise have resisted. Left clamped for a long time, that same foam cushion also continues to slowly flatten and settle through ordinary creep even though nobody has touched the clamp screw again, gradually losing thickness and, with it, gradually loosening the clamp's actual squeeze on whatever it was holding. Switching that foam cushion for a slightly denser, firmer foam changes how far the clamp screw has to turn to reach the same clamping force but does nothing to stop the same slow, creeping settle over time, since a firmer foam still creeps, simply more slowly and by a smaller total amount than the softer one would over an identical stretch of time. A gasket clamped inside a bolted joint behaves exactly like that foam cushion in both respects, absorbing far more of the joint's compression than a solid metal-to-metal contact ever would, and continuing to creep and settle slowly over time in a way that steadily bleeds away preload the bolt was never given any way to replace on its own.
Why this creep problem never shows up in an all-metal joint
An ordinary metal-to-metal joint barely creeps at all under normal working temperatures, since metal's own resistance to slow, sustained deformation under load is high enough that the joint's clamping force, once correctly set, stays close to its original value for a very long service life without any further attention. A gasket material, chosen for softness and sealing ability rather than for resisting slow deformation, creeps far more readily under the same sustained compressive load, which means a gasketed joint's preload genuinely does decay over time in a way a well-designed all-metal joint's simply does not, turning "tighten it once correctly" into "tighten it correctly and then check it again later" for any joint relying on a compressible gasket. Raising the operating temperature makes this worse for almost any gasket material, since most materials creep considerably faster when warm than when cold, which is exactly why a gasketed joint operating near an engine or a furnace needs its re-torquing schedule set tighter than an identical gasket sitting at room temperature would ever require.
The one number worth remembering
A gasketed joint can lose a meaningful share of its original clamping force purely to gasket creep within the first period of service after initial tightening, well before any external load or vibration has done anything to the joint at all, which is exactly why gasketed joints so often specify a deliberate re-torquing step some time after initial assembly, a step an all-metal joint relying purely on the arithmetic covered earlier in this set essentially never needs. Most of that creep tends to happen early, in the first hours and days after tightening, which is exactly why the re-torquing step is scheduled once, a short while after assembly, rather than left as an ongoing routine maintenance task repeated indefinitely.
Why this matters in practice
Designing or maintaining a gasketed joint means treating the gasket's own compliance and creep as first-class parts of the joint's behaviour rather than as a minor complication layered on top of the ordinary bolted-joint picture, choosing a gasket material with creep properties suited to the joint's actual service temperature and duration, and scheduling a re-check of preload after the gasket has had time to complete most of its initial settling rather than assuming the first tightening will still be accurate months later. Skipping that re-check on a joint that genuinely needs it is one of the more common, quietly preventable ways a gasketed joint ends up leaking or loosening well before anything else about the design was actually at fault, an oversight easy to make specifically because nothing about the joint looked wrong on the day it was first assembled.
Where this stops being true
A gasket chosen specifically for low creep, a solid metal or metal-composite gasket rather than a soft compressible one, recovers much of the all-metal joint's original behaviour, resisting sustained deformation almost as well as the surrounding clamped material and needing far less of the deliberate re-checking a genuinely soft, creep-prone gasket demands. The concerns in this article scale directly with how soft and how creep-prone the specific gasket material actually is, not with the mere presence of a gasket as a category, and choosing between a soft, easily sealing gasket and a stiffer, better-behaved one is itself a genuine design trade rather than a choice with an obviously correct answer in every case.