Steam locomotive function: Big Boy

     
New explain video

Steam locomotive function: Big Boy on Youtube
with many basic explanations but also some type specific details

If the above URL does not work, please use this one.


[Hide]

Why new balancing concepts?

     
Recent status: Limited maximum speed

In times when the basic structure of steam locomotives and the according solution for mass balancing was designed, high speed was not relevant.
So mass balancing was solved as simple as possible, even without any additional parts.

Show balancing basics diagram

Image source: https://trumpetb.net/loco/cbal.html

Since then, the geometric structure of mass balancing has not changed. Only dimensions and masses had been optimized within the existing structure.
The finally abandoned 5AT project (more) proposed lightweight materials and design, same as large wheel diameters and short piston stroke - but nothing structurally new. A maximum speed of 200 km/h was expected there.

To meet today's requirements on main lines and give modern steam locomotive traction a real future, this limit definitely must be overcome.
Therefore a new solution for mass balancing must be applied.

The following sections explain some promising approaches and their backgrounds.


[Hide]

Practicable qualitative balancing for "two cylinder" engines

     
Significantly more maximum speed

A new concept of counterbalancing enables steam locomotives to run significantly faster than before, without needing more than 2 cylinders.
There is an ancient found impracticable solution with counter-rotating balance weights springing along with the main axle against the main frame, following the same principle like in IC engines.
Our new concept avoids the disadvantages of this solution and offers a solution much cheaper than one or two additional steam engines within the main frame:

---

Regarding mass balancing of the reciprocating parts in steam locomotives, I only had the idea of arranging two balancer shafts ideally 180° opposite each other around the driving axle, as in internal combustion engines in motorcycles e.g. with one cylinder or with two "twin" cylinders. However, in order to prevent alternating axle pressure on the track (which would be the main purpose), these would have to be sprung together with the driving axle, so the mass inertia of the driving axle would be more than doubled with a corresponding frame, which is probably rather unacceptable.
Recently, I came up with a solution that tended to be better: The balancing masses must rotate coaxially to the driving axle, i.e. the most practical way (but not the least mass) is to be mounted on the driving axle within the frame and e.g. as with the intermediate wheels in differential gears (bevel gears or pairs of spur gears ) are driven by the driving axle.
However, this would still almost double the vertical mass inertia.

Now I came up with a much better approach:
Due to the lack of sprung crankshaft suspension, the most obvious solution when balancing masses in internal combustion engines is always to overcompensate the crankshaft with regard to the rotating parts, because this minimizes the effort required for the balancing shaft technology. This overcompensation is the only reason why a balancer shaft sprung with the crankshaft is necessary at all when adopting this solution on steam locomotives.
With steam locomotives, however, the right solution would be not to overcompensate the rotating parts. This eliminates the need for compensating components sprung with the crankshaft.
What remains to be done is simply to solve the balancing of the reciprocating parts separately. And this can be done with 2 balancing shafts running in opposite directions in the locomotive frame, as long as the axle bearings in the locomotive frame do not have too much end play in longitudinal direction. The transmission of movement from the drive axle or a coupled axle to the first balancer shaft can be done by means of a chain or two coupling rods in the longitudinal direction (or three, and a concentrically mounted coupling rod on the other side), and the transmission from the first to the second balancer shaft simply by means of a pair of gears. If you want to have it completely torque-free and keep the motion transmission exactly horizontal, there would have to be 3 balancer shafts, a small one above and a small one below the middle one, which is twice as unbalanced and driven by an axle. But you don't have to be that precise, even at high speeds.
The offset effects between the left and right steam engine shouldn't play a role in the description here (but of course in the dimensional layout), if I haven't overlooked anything.


Get in contact and see what is possible!


[Hide]

Additional balancing of valve piston

     
More than best main drive balancing

A probably new concept of counterbalancing for the valve piston enables more cylinder power and even more maximum speed than the best main drive balancing alone.

The concept works as follows:

• A counterweight is moved by a bar and rods, coming from the valve stem, in opposite direction of the valve piston.
  So far this is nothing exciting, but let's connect this with another aspect:


• The 5AT project intended to use 2 piston valves per cylinder which travel exactly parallel, driven by the combination lever.
  The advantage of two piston valves is that the ratio between frontal area and circumference of a valve piston declines when the diameter declines, so it is useful to have more valve pistons per cylinder, to have a lower sum of forces on the valve stem(s) in the moment of exhaust opening if the sum of circumferences and with it the port opening area remains the same. Or to have the same force as before when doubling the sum of the port opening area. Or between these two options.


• Now connect these two above points: The valve gear drives one valve piston which is connected to another valve piston (instead of a "dumb" counterweight) as described above. The second valve piston has the opposite admission structure of the first valve piston, e.g. the first one has inside admission and the second one has outside admission.


• So far, the valve gear does not need to be stronger than the one of the 5AT project. But as the valve pistons' movement is balanced, we may rise their weights by using double admission types for each of them. This way we get four times the port opening area with the same valve piston frontal area as a "normal" single valve piston per cylinder. For that, we only needed a little more strength of the valve gear parts.

Get in contact and see what is possible!


[Hide]

Perfecting ACE 3000 balancing

     
More than best main drive balancing

The ACE3000 project is described in The Ultimate Steam Page by Hugh Odom.
It was a really promising project at its time, but then oil prices declined and decisions in the short term were made.
More to be read on the above page.

Show preview picture

Image source: https://www.american-rails.com/ace-3000.html

Here I will go into more detail on the aspect of mass balancing.

Show design drawings

Image source: http://www.trainweb.org/tusp/ace_det.html
Inside connecting rods are used to tie the two engine units together 180° out from each other.


Image source: http://www.trainweb.org/tusp/ace_det.html
(Click to open in a new tab or window in full size.)
The following text passages refers to the valve gear layout of this drawing.
Please note that the combination lever bearing points of the right valve gear (let's call it front valve gear) are oriented for inside admission and the combination lever bearing points of the left valve gear (let's call it rear valve gear) are oriented for outside admission.
The drive axles are mounted in same orientation referring to the orientation of the cylinders each, but not referring to the whole loco. So all radius bars were moved below the expansion link bearing when driving forward, and vice versa.

---

Some basics of mass balancing on steam locos and additionally some thoughts about the application of balance shafts:

Counterweights on drive wheels are mostly used to balance the rotating parts and a bit of the reciprocating parts, depending on the maximum speed of the loco. All of the counterweights that is more than only for the rotating parts effects an alternating vertical force on the drive wheel. If half of the reciprocating parts are balanced by the counterweights, the resulting force ist about constant (in terms of amount) and rotating in opposite direction of the drive wheel. So if a balance shaft is applied with a counterweight that is directed opposite to this force, it can compensate that force, and the engine is, concerning only the resulting forces (but not torque), balanced for its rotating rpm. (But still not for higher frequencies resulting from the limited length of the connecting rod.)

If the above balance shaft is not concentric with the driving wheel (or to be exact about the influence of the main rod gravity center horizontal offset: If it does not have a certain small horizontal distance from driving wheel which but is not practically feasible), the resulting force of the driving wheel and of the balance shaft, along with the distance between them both, results in an additional alternating torque that tends to rotate the engine. This can be prevented by using two balance shafts instead of one, and their correct positioning.

However, on steam locos these balance shafts would have to transmit their forces directly to the drive wheel and not to the main frame as soon as the drive wheel is sprung. This results in a much bigger inert mass of the drive wheel axle, which results in damage of the drive wheel or of the bearings when running on uneven tracks. I had read that only the higher stiffness of Boxpok wheels already led to bearing damages.
So this solution is not suitable. (A much more suitable solution is to be found via the above link "Practicable qualitative balancing" below "Recommended concepts".)


Regarding the potential of the ACE 3000 concept:

At the ACE 3000, the two front drive axles are connected to the two rear drive axles with an angular offset of 180 degrees.
So the reciproating masses of the front piston etc. and of the rear piston etc. can (if having the same weights) compensate each other concerning their inertia forces in longitudinal direction. This not only for rotating frequency, but also above.
So there only the rotating parts (and some part of each main rod) need to be balanced by the counterweights in the drive wheels.
A really superior solution, but there are even more options for balancing:

If e.g. the rear eccentric crank would also be oriented trailing but referred to loco forward direction, then also the front and rear eccentric rods would compensate each other in longitudinal direction (concerning rotating frequency, but not exactly above that).
Perhaps the mass of eccentric rods was too negligible to make different drive axles for front and rear?
Although an horizontal eccentric rod layout would have been enough to just mount the rear drive axle the other way around instead of having different dimensions.
Or was the lower expansion link bearing stress preferred over a (little) better balancing, when driving forward?

If all combination levers and valve pistons etc. were made for same admission type, preferably inside admission, also the front and rear valve piston would compensate each other concerning mass inertia.
I do not see why the rear combination lever and valve piston are made for outside admission, if not because of the higher position of the rear valve piston.

So there seemingly could be an even much more complete mass balancing than only for the main drive, with no extra cost. Not for higher frequencies (than rotating frequency) except for the main drive, but much more than any real steam locomotive has had until now.

This could make this locomotive concept highly competitive also for fast passenger trains.
At least if the last really big disadvantage will be solved - the startup time of the boiler.
The good news is that this already is solved - for more, see Mackwell - Boiler.



Need some fresh thoughts and points of view from "out of the box"?
Get in contact and see what is possible!


[Hide]

Going beyond ACE 3000 balancing

     
The best practical overall balancing

The ACE 3000 project (see the above box) is a good basis for practical balancing, but there is a lot more that can be done practically to reach maximum speed and meet today's mainline requirements.
For the following explanations, a common basis is the connection between front and rear engine which ensures a 180° phase offset between front and rear engine. Today we recommend cardan shafts and angular gear boxes for this.

(For understanding the following explanations, you need knowledge of the names of valve gear parts. These are to be found here.)

Show valve gear part names

Screenshot source: https://en.wikipedia.org/wiki/Walschaerts_valve_gear#Technical_details

First variant of modifications:

• The main drive ist balanced not only for rotating frequency, but also above. So this must be maintained.


• The valve piston movement is not balanced, but this would be easily feasible by changing the rear valve gear to inside admission.


• Doing this, the orientation of the rear eccentric crank still does not balance the front eccentric crank and eccentric rod, so this must be changed.


• Now the rear eccentric rod and valve piston move in opposite direction, compared to the front engine set.

What remains is the partial vertical movement of the main rods and the eccentric rods in opposite direction to each other rear and front. These can be partially compensated via counterweight correction on the wheels. Full compensation (only with rotation frequency) would be possible using balance shafts, but as long as full compensation is not necessary regarding the target max speed, this solution cannot be justified.


Second variant of modifications:

• The main drive has its cylinders directed each one to another, so the distance between the 2nd and 3rd drive axle is much larger than at the ACE 3000. But we can use the PRR T1 as a prototype, where one set of cylinders was placed between the 2nd and 3rd drive axle, and provided the same fixed axle set length, a wheel diameter of 72" instead of 80" at the T1 should be far good enough, as inertia forces increase only with wheel diameter decreasing to the power of one.


• If allowed by max speed, the second valve gear can be omitted and the first valve gear controls all cylinders using a prolonged valve piston.


• If this applies, our new version of valve gear can be applied. From today's point of view, it requires a leading eccentric crank, related to cylinder forward direction.


• If max speed does not allow to omit the second valve gear, it can be used for only moving a counterweight instead of a valve piston, or separate valve gears with valve pistons can be applied as described above.


• If max speed does not require a second valve gear but at least a linear moving counterweight for the valve piston, a separate valve piston for the rear cylinder can be actuated via a return lever driven by the regular valve gear, or if two valve pistons per cylinder are wanted anyway, the second valve piston for both cylinders can be driven this way and have an opposite structure to the first valve piston, e.g. with outside admission for the first cylinder if the first valve piston has inside admission for the first cylinder.

Some thoughts about including the partial inertia of the main rod in the balancing calculations:
In side view, a material part can be represented by two mass points. Calculating these for a main rod, their position will be neither on a bearing point nor on its gravity center.
A representation which allows easier application of masses to bearing points would be with 3 mass points, one of them applicable on the gravity center of the mainrod and both others applicable on the bearings.
(I am not quite sure but I think that this representation also is applicable not only in side view but also in full 3D concerns, except for the fact that the first point on non-symmetrical main parts does not have to be on the position of the gravity center but can be used to adjust the gravity center. I hope that there is no error of thoughts in this.)
Based on this representation, the mass points on the bearings simply can be attributed to the wheel and to the crosshead, whereas the resulting inertia forces of the mass point on the gravity center of the main rod can be converted and attributed partially to both of these bearings.
Beneath easily compensable effects, this shows that mass forces result in a partial vertical inertia force on the crosshead bearing which only can be compensated by something that is not being sprung like the drive axles. This e.g. could be a balancing shaft (compensated in horizontal direction by a balance shaft on the opposite end of the loco) or a set of balance shafts.
For our SAS 3000 concept, we assume that the latter solution will not be necessary, regarding max speed up to 280 km/h or a little above that.


Beneath balancing for high speed, here some more other conceptual aspects:
The engines should be kept as simple as possible to avoid the lack of practicability of the ACE 3000 project seen by L. D. Porta, so we recommend single expansion (no compound system). Reaching out for carbon neutrality, this should be really acceptable from the point of overall lifetime costs.
For heating up quickly and for being able to burn materials named as carbon neutral, we strongly recommend using a Mackwell boiler. If necessary, a second boiler can be placed in a rear sectional part of the locomotive, looking similar to the ACE 3000 rear part with small running wheels.
As there is no risk of explosion with the Mackwell boiler, automated control of the second boiler should not result in any bureaucracy.


To be continued... (and possibly corrected; current version uploaded on 2024-10-18 18:50 CET)


Need some fresh thoughts and points of view from "out of the box"?
Get in contact and see what is possible!


[Hide]