Steam & Excursion > Heat treatment ovens and stress-releaving steel frames.

Back in the day, RRers were often confronted with broken loco frames,
very similar what is the case with 4070.

In reality, successful weld repairs were rare, after the engines were 
back in service, hauling freight.   The 'repair' simply reappeared, when 
faced with heavy trains.

The early one-piece, frame products of General Steel Castings
suffered cracking problems in the area between the cylinders and the 
first driving axle.  

Decades ago, when I was in Bellows Falls,Vt, and Blount's Steamtown,
I met a retired worker who had worked at General Steell Castings, and he 
told me of the problems with a lot of their earlier. one-piece frame castings.
( cylinders cast as part of the frame) for locomotives.

He pointed out an example in the Museum's Collection:
4-8-2, GTW # 6039 ( 6/25) has a broken frame condition at 
the right-side of the cylinders and frame.  Sure enough, there is
the cracked and broken cylinder casting at the frame, with evidence of
extensive welding repairs that were used to try repair the original cracked
areas.

The existing cracks are a testament to their failed attempts at welding
repairs that were apparently tried & used several times on #6039.
Go, and see for yourself, next time you're in Steamtown.

The guy from General Steel Castings told me of their later successes 
with the castings resulted from an almost 100% increase in size and
steel added in their later-production cast, one-piece frames.
What he pointed out to me was the much more massive curves -- between
the cylinder castings and the long frame rails.  The greater, stronger
proportions can be found on locos like the NKP 759, Berkshires.

So, let's examine the 'physics' of forces involved....

One of Newton's Laws of Motion states that every action has an opposite
and equal reaction.  

With pistons in cylinders in steam locos, the pressure in the cylinder presses in
all directions, and "it's equal and opposite forces" are the pressure exerted on
the adjacent cylinder head.  In terms of tractive effort, you 70,000 pounds acting
on the piston, you also have 70,000 lbs acting on the adjacent cylinder head.

When starting, and the right-side piston is moving, the crankpin and 
wheel will be moving from the 3 o'clock position, down through 6 o'clock,
to the end of stroke, near 9 o'clock postion.  

If the driver does not slip, then, to move the train, the steam in the cylinder
pushes against the front cylinder-head ( with full tractive effort) & moves the frame,
and pulls on the drawbar connecting the tender & train.  

Thus, the frame takes a lot of force in getting things underway.  What the builders
didn't fully appreciate was that the forces of "equal and opposite" pressures are
always at work.

If we were next to a starting loco & train, and we had made a cut in the frame,
as a visual aid, above #1 drive axle, and we could watch the crack, we could see
the cylinder-head pressure as it opens our 'crack', in pulling the train.

With the piston at the end of its out-ward stroke, steam will be admitted to back 
of the piston, and while the crankpin is over the top of its curve, our
'visible witness crack' will close-together,  still pulling against our cars...

The steam pressure still presses on BOTH the rear cylinder head and the rear 
face of our piston --- with equal pressure--- as it moves forward towards the
front cylinder head.

The classic method of preparing to do any extensive amount of frame welding,
using stick-welding techniques, was to build an "oven" of refractory brick
( at the cracked area) on all sides, packed with charcoal, and set 'cooking'
for a couple of days.  

Then, a wall or two of the oven would opened to allow a weldor to start a
short repair ( in short time-frame) area where he would add metal.

Welding-Arc temperatures are a couple of thousand degrees, so great
temperature differences occurs between the base-metal ( frame, typically)
and the added metal of the welding sticks.  

After a brief welding period, they close up the firebrick oven, add fresh charcoal,
and let the temperature increase inside the enclosed 'oven'.

To make the repair, you're applying a lot of steel, by weight ... so short 
welding times are crtitcal to try to limit the total 'thermal expansion' that 
occurs at arc-welding temperatures ---- of thousands of degrees...
Plus you're laying-in a heavy amount of new metal.... and aa intended continuous welding method soon has greater thermal 'growth' than the intended amount.

You have also built a large volume of
higher density steel that creates
fault-line between the greater density welded-in mass of new steel
that is potentially a starting point
for new cracking to occur.

So, you can see how large amounts of force are applied by the two pistons,
in continuous, power-cycles that are strong --- in two directions.

Now, let's go back to our imaginary slit that opened and closed with each
360-degree cycle of two power strokes.  The gap is closed, but because of
possible under-engineering, the strong forces are greatest at the joint, right at
the cylinders, and the frame, before the first driver set.

As mentioned earlier,  because of inadequate strength ( by poor design)
at the connection between the cylinders and the frame, it's the most common
location for frame breakage.

Simply welding-back to repair the crack is still inadequate compared to the 
forces applied during powering-up.   The cracking area needs to be buttressed 
with steel plates to be added.... to strengthen that whole area.

( Added plates and Huck bolts, sandwhich style, could be a workable alternative...
   Be sure to upgrade both sides of the loco...
   But you'd still need to analyze the location where the cylinder casting
   matches to the frame..)

Welding is tricky because of the extreme differences in temperatures of the
welding arc, and the broken frames... even when heated.
It's an unavoidable circumstance that is commercially resolved by using immense
heat-treating 'furnaces' and long 'cooking' time to 'normalize' the weldment structure.

 In repairing a broken frame, ideally you'd want to do small amounts of
welding at a time, and using a refractory-brick oven; then, finally a trip to the
'normalizing oven', for a week's "heat treatment": 20%  heated-oven time,
and 80%, a week's long cool-down period --- with a closed oven & no fire....

Even after all of that, you will still be faced with the inadequate basic, frame design.
Your welding job might be excellent, so the next crack will be in the 
adjacent area of the same weak frame.... possibly adjacent to your recently 
welded area, that stood up to the forces, .... but, the weak frame is still a weak frame;
it can always 'give' at somewhere else...

So, the benefit of attempting a welding repair in that area is most useful as
an educational opportunity for repairmen and weldors.

But, to solve the weak-frame design means a whole engineering approach 
( to a patch) that strengthens and reinforces both sides of the frame: 
right frame-rail & left-frame rail  and cylinders.

A really strong right-side repair, simply means that the cracking will show-up,
across, on the opposite cylinder and the same weak fame members..
as on side 1.

Just because you fix one side doesn't mean that the problem goes away.
And of course, since the cylinder events combined with weighty-trains
is the true cause of high pressures at the pistons,  you can limit piston 
forces by only hauling considerably lighter trains..... probably under 9 cars,
wiould limit cylinder pressures, so that you never fully-force the pistons to
deliver their full, intended power.

​W.

( Lowered piston forces is the whole reason that PRR tried to go with duplex 
  drives, like the ill-fated T-1class 4-4-4-4 locos. Their smaller-diameter piston forces were only
driving 2-main driving axles... and you had 2 sets of smaller pistons & cylinders and
powered the drivers, with much lower piston thrusts, than comparable 4-8-4s).

( Not proofed, yet...) 



Edited 21 time(s). Last edit at 12/23/24 23:36 by wcamp1472.

could you huck bolt these things?

I've read this article in Classic Trains and borrowed a photo from it.
I stare at this new front cast frame and wonder how this could  be welded to an existing frame (after removing the old frame at this point) and since it is welded at the pedestal opening of the #1 driver not be subject to cracking problems.  These were big and very powerful locomotives and the forces have to be very high.
I realize the binder does it's job, but this baffles me how they hung together.
The preheating, welding and stress relieving must have been very carefully done.
I've been down to Continental Fab here in St Louis when Scott Lindsay or Gary Bensman were in town reviewing boiler work being done and was intrigued by their oven, which would hold the whole boiler for heat treat. Very impressive.

---

Re: the new partial frame in the photo.
       This partial frame pre-dates the final development
        of the cast, one-piece, frames & cylinders, used on later locos.

Note the very substantial 'binder-bar' at the very bottom of the frame opening 
for the #1 axle and the journal bearings.  They are on either side of each axle.

There is a critical, small space between the top of tte binder-bar and the lower 
edge of the frame.  The binder ends have tapered, wedge-like frame braces
that match the tapered protrusions of the frame.  

The 'binder bolts' have large, 6-sided nuts and lock washers.  
After nearly every trip some of those 'binder bolts' will have worked 
loose, and need to be re-tightened --- best done while over an inspection 
pit.  You always want that 'wear-gap' to be visible and measurable...
When the "wear-gap" disappears,  the frame can be moved, ever so slightly,
but that condition will worsen..   You want the binder-bars tight against their
wedge frame member. 
 
Eventually, the wear is enough that the entire binder is worn, and that 
visible space gets too small and the binder 'bottoms' against the frame!
A condemnable condition, and the engine removed from service until
replacement binders, with proper tapers, are available.

Smart shop supervisors, always had spare binder bars, with wedge faces
machined for various amounts of wear, and ready for replacement, as needed.

That small gap is crucial to all frame binders and binder bolts.
Wise tourist RRs will hammer-rest those binder bolts after every trip,
or after each day's train hauling duties.

​W 



Edited 2 time(s). Last edit at 12/26/24 08:18 by wcamp1472.

I'm aware of what you said about the binders.  I'm still curious as to how that new frame portion can be secured to the rest of the frame strongly.  The heavy binder would do the job on the bottom of the frame.  The top rails are angled to a point and I wonder if the remaining frame rail had the matching angle or it is for doing a heavy rood pass and additional welding to make the joint.  No matter how, it's going to have to be one helluva good joint.
FWIW, in the whole time 1522 ran, the binders stayed tight and we didn't have to take them up.  Of course she wasn't in heavy daily service either.

"Re: the new partial frame in the photo.
       This partial frame pre-dates the final development
        of the cast, one-piece, frames & cylinders, used on later locos."

Photo is 1944, and the pictured cast front section is for rebuilding/upgrading some very large GN 2-8-2s.
By 1944, complete cast engine beds had been successfully produced for over 15 years.
Interesting factoid:  The "reject rate" on the cast engine beds approached 50%-- quite a few castings went back into the GSC melt pot!

Very interesting explanation of the forces on steel frames.

I remember being shown the crack in the cylinder saddle of SP/TNO 786 in Austin when it was first discovered in 1999 and yes, one could see old welding beads where repairs had been made to fix the crack. 

Would something like the Thermite Welding they do with CWR work on a cast frame?

Thermite welding is an excellent process….. However, it’s the frame that’s weak and
under-designed amount of frame steel.

Under heavy loading and high pressures on the pistons, any new weld is now very strong,
so experience has shown that a fresh crack will appear right next to the welded area,
in the original material that keeps cracking … but, again, it’s the over-stressed, inadequate
amount of strong, reinforced steel at the right-angle joint where the cylinder casting is
mounted to the frame.

It’s the constant, alternating pressures in the cylinders that causes the weakened steel
to crack and fail.

It’s interesting that the photo of the GN, 2-8-2, loco new frame section exactly addresses
the type of extensive re-do that is required to to solve their problem.

The original design, soon was failing—- even though the engines, at tte time, had the
greatest piston forces of any 2-8-2. But piston forces are only half of your design challenge…
If the frame is comparatively weak and inadequate in strength, by poor design, that’s the
true cause of the continuous frame cracking problem.

Fixing that whole area of a locomotive frame, as in the 1941 photo, shows the extent
of the re-engineering that is required to solve the frame-cracking issue.

No amount of re-welding and repairs will solve the bigger issues of frame cracking
problems.

Fresh-breaking, through the newly welded frame repairs, is an indication of inadequate
preheating and pre-welding temperatures are too low inside the firebrick-and charcoal
heating oven..

All newly applied welding material is very brittle, compared to the base-metal of the frame . …
.and the welded-area cracks very easily, if not properly 'normalized' in a suitable 
heating oven --- cool-down after hours at the proper high-temperature, takes  5 to 7 days..

In service, the weld-repair area may hold-up well;  but, the cracks will continue to appear in
'fresh areas' of the original, weak frame structure.

W.

Posted from iPhone



Edited 4 time(s). Last edit at 12/26/24 07:44 by wcamp1472.

Wes:

With 20% at temperature time and 80% cooling time, you’re describing an “annealing” process, not a “normalizing” process. While in this case annealing may be the best method to reduce the possibility of recurrence, normalizing generally involves soaking the part at its critical temperature for the proper amount of time, followed by an atmosphere or even air “quench”. The microstructure between the original and newly added steel is combined/reformed into even-sized “blended” grains, thus the term, “normalized”, but the toughness of the steel remains. Annealing also results in a reformed, blended grain structure. However, if cooled slow enough, as suggested by the 20% at temperature/80% cooling time frame, the grains of steel can take on a rounded appearance under the microscope. This is known as “spherodize annealing” which results in an extremely soft part or in this case, portion of the frame.

That said, what material was a typical cast locomotive frame made from? If the percentage of alloy and carbon were high enough (52100 for example), the cooling time wouldn’t make much of a difference. PLEASE KNOW - I’m NOT “knit picking”. Simply curious and very willing to learn.

Mark V.
“The Fort” in Indiana

Posted from iPhone

I appreciate the distinctiion that you describe.   Refining the terms 
helps everyone get a better understanding.  Thanks for your essay.

I have looked at tte crack on 4070, when I visited the restoration 2100,  
at a dinner they had a year or two ago.  It didn't look like a cast frame.  
Yes, the cylinders are castings,  a right and left halves are bolted together.

But, the frame was not a cast, one-piece frame, that would not have been available
at the time.  I think the frame was 'built-up' of large sections of plate steel.

The loco was built in 1918, about 10 years before cast, one piece frames were 
introduced.  And I doubt that 4070 was later equipped with a new cast, one piece 
steel frame.  And, yes the eariy cast, one-piece frames with integral cylinders were
plagued with cracking frames in that area,  beacuse of inadequate proportioned
angular-bracing between the cylinders and the joint with the frame rails.

This seems to be case of inadequate designed proportions needed at this 
critical structural area.    But, at the time it was designed, the proportions 
were logical, considering the light train-weights of those times.

2-8-2s, in 1918 were considered a large step forward, with 2-8-0s being the
former 'standard' freight engines.  So, designers were guessing about dimensions
and the magnitude of forces at work, in hauling heavier and heavier trains.

An area that was poorly understood was the effect, over time, that 
continuous, very strong reciprocating forces ( continuous back-and-forth pounding
and constantly fluctuating power strokes' intensities).  
The strong repetitive cycling, with hauling heavier trains was not anticipated at the
time, in 1917, when this loco was being drawn-up for construction.

A factor that also was not fully appreciated in 1918, was the impact of 
improving Westinghouse freight train air brakes and cross-compound
air compressors.  The, common, smaller freight train cars and components
were growing at faster rates of improvement, and were being built to carry
heavier and heavier carload-weights, and fitted with better car air brakes.

A dramatic increase in capacity and reliability of freight-car brakes arrived in
1933, with Westinghouse introduced the "AB" brake valves.   That system led
to a whole new approach to freight-car capabilities, capacities and newer designs....
far exceeding the standard freight cars & trains of 1915-era.  

It's no wonder that 4070's frame broke --- when placed at the head of trains 3 and
4-times the weights anticipated in 1917.

When 4070 was on the drawing boards, no one could have anticipated the coming
 impact of better and better automatic airbrakes, and the very successful #6ET 
locomotive brakes, and it's improved engineer's brake valves, over the quaint,
but inadequate predecessors.  The better automatic brake valves, coupled with
the new AB car air brakes, could control brakes on trains 2 and 3 times more cars
than the earlier brake systems.

Progress in all aspectcts of steam locomotive technology exploded after WW1,
the era 4070 was designed and built.   In 1917, there were no available mass-produced,
successful, power coal-stokers when 4070 was ordered.  The traffic 4070 was designed-for
was rapidly, and dramatically, being increased in the coming decades,after 1918.

​W.

( not proofed, yet...)



Edited 4 time(s). Last edit at 12/27/24 16:52 by wcamp1472.

Question:
In May, 1977..... 2102 + 4070, double headed, made trips around Horseshoe Curve...
... 4070 failed and a bent an eccentric rod...

Which side of the 4070 had the bent eccentric rod, fireman’s side or engineer’s?

Had 4070's service crew known about the broken frame at the 
right side cylinders, at the time of, or before they ran those trips?

​W.


 



Edited 5 time(s). Last edit at 12/28/24 09:07 by wcamp1472.