Rivets or Fitted Bolts.. Ex FS thread.

Interesting one is this. If you look at any Gresley designed loco, you will see that, on the frame structure, fitted bolts are used extensively as fastenings. Bulleid also used these on his Merchant Navies and to a lesser extent on the WC/BBs. Now, when I viewed Scotsman's frames a few years ago, it, too, resembled a piece of meccano with many fitted bolts used. From photos on the NRMs Scotsman blog it does not look like meccano now. Indeed, one of the short film clips shows a horn being hot riveted in place. Various engineers have decided on this course of action for its reconstruction. It should also be remembered that the frames have been inspected twice by respected engineers and reports done. So who am I to criticise.

 

I would have thought that hot rivetting is not the best way to fix hornblocks/guides as the rivets shrink on cooling and leave the connection dependant solely on friction grip. Fitted bolts, alternatively, completely fill the hole and place the bolt in single shear as well as the friction grip resistance.

 

So, if you knock the head off a rivet, it will just fall out.........................?

 

Steve, rivets are a friction grip fixing. If a head is knocked off, the friction grip is lost and any shear or tensile stress is re-distributed to the other rivets in the connection, overstressing them. Because rivets shrink on cooling, there is always a tiny annulus space around them and for them to take shear force, there must be some movement in the connection. For movement to take place in hornblocks for example, the other rivets will have been overstressed and the frictional resistance of them will not be enough to resist the small movement. Any movement in this scenario is considered a failure of the joint.

The rivet may not necessarily fall out for a number of reasons but nevertheless the damage is done. Fitted bolts are hammered in cold and as there is no annulus space they will take shear force without allowing movement.

It's all a bit theorectical but it's all based on rivets shrinking on cooling..

 

All theoretical, as you say. Bolted joints are a friction grip and, to get that friction grip, you have to apply a tensile load to the bolt by tightening the nut. Probably teaching you to suck eggs but that tensile load results in a very small reduction in diameter of the bolt and that applies whether it is a fitted bolt, or not. So, both fitted bolts and rivets suffer a very slight reduction in diameter when finally installed, unless you leave the fitted bolt slack. Whatever the theory, a well squeezed river it invariably tight in its hole and even the best fitted bolt joints can work loose. You pays your money and makes your choice. Perhaps the solution is a combination of rivets/fitted bolts and dowels!

 

But isn't the idea with fitted bolts that they're made several thou oversize, so any reduction in diameter by the stretching would presumably be taken up by the extra diameter?


Sent from my iPhone using Tapatalk

 

So how do you get the bolt into the hole in the first place? If it fits in the hole, then you torque it down, it will then 'stretch' and become undersize as said above?

Maybe we need to invent the inflatable bolt ?

 

A fitted bolt is made 0.002"/0.003" oversize on diameter and then hammered in. The same applies to a cold turned rivet, which would be turned over hydraulically.

 

Fitted bolts are a knock in fit into a reamed hole which guarantees accurate alignment between the two components a fit of 1thou on a 1” bolt will be about right the use of an anti scuffing paste aids assembly they do not shrink when tightened up as it is the thread that stretches. There were 120 fitted bolt in the horns riveting is a lot cheaper but fitted bolts were what was used originally.

 

Inflatable bolts - no problem, just pop them in the freezer the night before fixing LOL

 

Yes I understand the basic principle of a fitted bolt (even though I'm only a dumb chemical engineer ) but it was the point made above about the bolt stretching under tightening and so ending up with some play that was intriguing me. So now we seem to be saying that either a) the deformation of the bolt and/or frames is elastic and maintains the fit as the bolt is tightened and 'stretched' or b) the bolt shank itself doesn't stretch, only the exposed threaded portion (not sure how this can be?). What is the primary function of the fitted bolt - to maintain the lateral location of two components acting in shear, or to draw together those two components through compression? Surely it's the former as the friction of the bolt fit wouldn't guarantee the two components mating properly when torqued down? So is there any necessity for high torques on a fitted bolt, or do you just need a means of preventing the nut falling off (of which there are many) in which case there's no 'stretch' to worry about ? I think perhaps we ought to move this discussion to the MIC thread?

 

There's an internet cliche on the lines of "this thread is worthless without pictures".
In this case we have "This subthread is worthless without numbers".
Unless supplied with actual numbers for the amount of deformation of rivets and fitted bolts, including also the actual statistical incidence of less than perfect installation, than all we have is a particularly useless. maybe even harmful kind of speculation. I don't doubt that engineering data on the pros and cons of different fastenings (which would include the effects of any wear too) is available at a sufficiently high level in the mechanical engineering industry, and I suspect that I wouldn't have the maths to understand the data properly even if I saw it...

 

Hmm. Not being a rivet-counter myself I am not unduly alarmed, but aren't they meant to be preserving / restoring it (as opposed to just 'fixing it up a bit') and if so, are you saying they have deviated from the original design? What next, German-style smoke deflectors and apple green... oh, hang on.

 

With regard to fitted bolts, I suspect they are designed to resist single shear and not primarily for friction grip. If one wants to use them for the latter as well there has to be a method to ensure that they are properly tightened up to develop the friction. In structural steelwork, torque wrenches are a no-no. This is because some of the torque could be used up when the holes don't line up, or if the nut is cross-threaded. In fitted bolts some of the torque would definitely be used just pulling the bolt through the metal and the actual friction grip could not be accurately acsertained. One way, like structural steelwork would be to have indicator pimples on the back of the head of the bolt so that the actual friction grip could be checked using feeler-gauges..

 

Whatever the rights and wrongs of fitted bolts v rivets, it is the bolt that stretches between the head and the nut and progressively less so throughout the length of the nut (assuming a perfect thread, that is.)

 

All this talk about HSFG bolts and friction joints leads me to think that you are involved with structural steelwork where you tend to design for a single direction load(?) The trouble is that bolts on moving structures (and a set of loco frames is a good example) seldom have a single axial or radial force applied. It is a combination of the two so some pre-load is necessary and, unless that pre-load is throughout the length of the bolt and of sufficient value, a fatigue situation will be present.
At the end of the day, the method of fastening is at the whim and fancy of the designer.

I only got involved with this discussion because, in my experience, a well fitted rivet is tight in the hole and takes some knocking out!

 

I think we all would agree departure from the designers "whim and fancy" is fraught with peril. I'm sure Sir Nigel and his team used methods that they had confidence in in preference to alternatives so why fix what ain't broke.
If FS has had its horn fixing method altered, as suggested earlier, I would think it incumbent upon whoever changed it to satisfy the appropriate approval authority that a satisfactory union has been made between the two components.
If true, is it documented who changed the fixing method?

 

It's many years since I designed any structural steelwork but the loads were generally multiple and often dynamic. I think contemporary practice generally uses computerised finite element analysis and often has to consider blast and earthquake loads as well as the traditional wind loads and rolling loads.

 

I suspect this was cost driven. The rivets can be replaced in modest batches when the locomotive has seen a few years service. Interestingly the late W O Bentley trained on the Great Northern and used fitted bolts on his cars. He did not like rivets.

 

I fear that modern loco engineers are empiricists by birth. The thought of using computer finite element analysis to design hornblock/frame connections would be anathema to them..