Trains.com

Radial Truck Development at EMD

20810 views
43 replies
1 rating 2 rating 3 rating 4 rating 5 rating
BDA
  • Member since
    April 2018
  • 95 posts
Posted by BDA on Friday, June 19, 2020 6:25 AM

I had a close look under the 90s here and the fuel tanks appeared to be seam welded for their entire length to the under frame . They have gusset plates and in some places box sections welded towards the tanks ends to the frame .

For some reason they'd occassionally get this reasonably violent cyclic thudding which I equate to earthquake mode . It felt like a really rapid loading and unloading of traction power and everything in the isolated cab would jump and dance all over the place . I had thought these issues were because of the early Siemens control systems and inverter antics .

70ACes don't seem to have any of these issues but they would be have the benefit of further development and improved electronics . 

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Wednesday, June 17, 2020 8:38 PM

Earthquake mode as I understood (I wasn't involved with trucks at that point) was a dynamic response to the frequency of inverter changing TE that corresponded to the bounce frequency of the wheelset on the motor nose link at about 6 Hz. Re-tuning the control system took care of the worst of it although at the adhesion limit it still happens.

The underframe cracking on 90MAC's was due to the very rigid mounting of the ends of the 6,000 gallon fuel tank to the underframe wherein the buff/drag load was carried into the fuel tank. Prior to the 80/90MAC, the fuel tank was held on long bolts passing thru holes in the heavy bottom plate of the UF to a heavy angle welded to the endplate of the fuel tank. With the longer, heavier tank on the 80/90MAC's and the desire of the manufacturing and service groups to have external fasteners holding the fuel tank to avoid having to remove and engine if the bolts needed replacing, the fuel tank had the angle welded about 6" lower on the end of the tank and a second angle with gussets between each bolt was welded to the bottom plate which resulted in a much stiffer connection. The first failures occurred as cracks in the fuel tank and they were campaigned to be opened up and significantly stiffened internally. This fixed the tank cracking but the stiffer tank led to the fatigue cracks in the underframe. IIRC, the bottom plate on the 80/90 is only 1" thick compared to the 70's and prior which ranged from 1.5-3.0 inches. The underframe is deeper on the 80/90MAC's which is why it worked with a thinner bottom plate but given the long length of those units the deflection at the center under 1M lbs. buff load is about 4" compared to the ends. The fix for the underframe cracking involved, of course, gouging out the cracks and welding them up, and then softening the mounting of the tank using bolts with spacers and spherical washers at one end and letting that end slide on the mounting so the tank didn't carry so much of the buff load.

Dave

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, June 17, 2020 6:15 PM

This thread could run for years and I would still be avid to read (and learn from) new posts.

bogie_engineer
The Phase 2 90MAC's were a bit different than other EMD locos with their integral fuel tanks that created a "diving board" effect which impacted the ride sensation.

Did this contribute to the frame cracking that put many of these units out of service... and were fixes to the 'effect' incorporated in the rebuilding (which seems to date to be reasonably successful)?

I'd be interested in the full official story of 'earthquake mode' (which I understood in the '90s to be in part a desire for 'more rapid loading than GE' combined with modal rather than gradual proportional engagement of traction control).  Since this was apparently characterized as observable on SD80MACs, was the 'diving board' excitation observed in other long-frame contemporary EMDs and not just the integral-tank variant?

It seems at least possible that earthquake-mode modulation within the period of vibration of the frame would easily contribute to some of the more exotic modes of isolated-cab motion.  I can easily see why vibration analysis, even in multiphysics modeling, might not predict this.

Is there an analogue to higher polar moment of inertia for the end axles of HTCR trucks, due to the lever system or any individual yaw-excursion damping of the axles, compared to the simpler dynamics where all three axles are 'normally' constrained in the less expensive version?  

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Wednesday, June 17, 2020 11:25 AM

The Phase 2 90MAC's were a bit different than other EMD locos with their integral fuel tanks that created a "diving board" effect which impacted the ride sensation. From a truck standpoint the vertical and lateral ride should be very similar since the suspension components are all the same but ride encompasses everything from the seat to the rails so it's not surprising they feel different.

Dave 

BDA
  • Member since
    April 2018
  • 95 posts
Posted by BDA on Wednesday, June 17, 2020 5:47 AM

Rode and seemed to track a bit better .

90s sort of rocked through corners and made a few strange noises . The 90s here are Phase 2s converted to 710 2 strokes , not sure what sort of mileage they would have done at UP with the H engine in them .  

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Tuesday, June 16, 2020 3:10 PM

The only reason EMD developed the HTSC-2 truck was to offer a non-steering truck as basic equipment at a reduced cost. Since the SD70 series debut, EMD did not have a NA size locomotive truck that wasn't radial steering which put EMD at a cost disadvantage against GE whose standard truck was the Hi-Ad and their steerable truck came in as an option, which due to it's cost and reliability, RR's did not end up buying in big numbers and those that did like CSX eventually gave up on it too. Both the AC and D90 DC traction motors would not fit in the previously standard HT-C trucks as well as the tendency of the HT-C bolster to tip on the center bearing under high adhesion made it not an option; not to mention by the time of the SD70ACe, the HTCR manufacturing cost was less than the HT-C due to it's complex frame casting and the extra cost of the bolster. When the SD70ACe program was conceived, it's primary design target was to reduce manufacturing cost, starting with changing inverter suppliers from Siemens to Mitsubishi. The HTSC-2 non-steering truck was essentially an HTCR truck without steering beams and primary yaw dampers; it uses the same bearing housings, primary and secondary springs, and vertical and secondary yaw dampers as the HTCR. Most customers did recognize the advantages of the HTCR trucks and continued to specify them. I do know that CSX was very upset that the first and only order of SD70ACe's they bought, the 20 prototypes, did not have HTCR trucks.

BDA

Hi Dave , Im interested to hear about the HTSC2 trucks under SD70ACes . I got to run on these and if anything they seemed to work better than HTCR2s under 90 MACs .

 

When you say better, what are you referring to, ride, tractive effort, etc.

Dave

BDA
  • Member since
    April 2018
  • 95 posts
Posted by BDA on Tuesday, June 16, 2020 7:11 AM

Hi Dave , Im interested to hear about the HTSC2 trucks under SD70ACes . I got to run on these and if anything they seemed to work better than HTCR2s under 90 MACs .

Cheers .

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Tuesday, April 21, 2020 12:15 PM

RMEX

Dear David,

Were you at all involved in the SAR 11E class e-lok bogie proposal? B-B + B-B proposed by EMD but the SAR decided to keep using the six-axle bogie having the traction-link bars and the ASEA traction motors?

Kind Regards,

BW Ring

 

I joined the truck design just as that project was nearing completion but so I wasn't involved in the HT-BB 4-axle articulated truck that was proposed by EMD as the alternative to the GSI 3-axle truck the RR much preferred. I was in charge of the test program on the HT-BB on the BN SDP45 6599 that we ran on BN's Stampede Pass in Sept 1984 so I know the design well. 

EMD did not have a truck that would accept the ASEA traction motors; the HT-BB was proposed because 8 EMD-size motors were needed for the 42" gauge to meet the tractive requirement. But Dr. Scheffel was not a fan of the HT-BB design and insisted on the GSI truck.

Dave

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Tuesday, April 21, 2020 12:07 PM

M636C

 

Dave,

From the photographs it appears that the GM10B had the "inverted swing hanger" arrangement on the secondary coil springs where the secondary springs were  attached to spring plank attached to the locomotive frame below axle level by swing arms. This allowed the centre truck to move sideays more easily on curves.

Peter

 

I haven't heard the "inverted swing hanger" description used before but as far as I know, the outboard trucks were similar in design to the AEM-7 trucks in which the bolster runs below the truck frame and has a pivot at the center connecting to the center of the truck frame so that the truck rotates about the bolster. The bolster has the coil springs on its ends that go up to the underframe supporting the loco. The bolster is also connected by an inclined, rubber bushed traction rod connecting to the underframe on each end which allows for lateral movement between the truck-bolster assembly to the underframe while it transfers the tractive force.
 
The unique part, to me at least, are 4 pendulum links that connect the boster to the truck frame that carry the vertical weightof the carbody in tension. These rods have semi-spherical pads on each end that allow the truck frame to rotate relative to the bolster for curving. This action creates a gravitational restoring force that centers the truck and helps with hunting stability.
 
The GM10B center truck was a similar design but had no rotational freedom between the bolster and truck frame. A really long traction rod was used to reduce its angularity when the truck moved laterally, I think there was about +/-10 inches of lateral travel. The end trucks had about +/-2 inches lateral freedom.
 
Dave
  • Member since
    April 2020
  • 1 posts
Posted by RMEX on Monday, April 20, 2020 7:39 PM

Dear David,

Were you at all involved in the SAR 11E class e-lok bogie proposal? B-B + B-B proposed by EMD but the SAR decided to keep using the six-axle bogie having the traction-link bars and the ASEA traction motors?

Kind Regards,

BW Ring

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Monday, April 20, 2020 9:00 AM

bogie_engineer

I was the noise engineer at the time the GM10B was designed so I did do noise testing on it but that was my only part of that project. The trucks and all the electrical gear including traction motors was an ASEA design. After the build and testing of the GM6C, which used modified HT-C trucks to accommodate the E88 motors with roller suspension bearings, EMD decided to do the GM10B for which the higher power required much bigger than the standard EMD motors. The ASEA motors used on the GM10B, and similar motors on the AEM-7 required 51" wheels so a whole new truck design was required. If EMD had designed the trucks, it no doubt in my mind would have been two 3-axle trucks but ASEA had the 2-axle truck design for their big motors so it was built in a B-B-B arrangement.

Dave

 

Dave,

From the photographs it appears that the GM10B had the "inverted swing hanger" arrangement on the secondary coil springs where the secondary springs were  attached to spring plank attached to the locomotive frame below axle level by swing arms. This allowed the centre truck to move sideays more easily on curves.

The Queensland Clyde ASEA Walkers freight locomotives had this arrangement.

My recollection is that the passenger locomotives (30 of the 80 were built with bogie frame supported motors and a higher top speed for passenger and intermodal trains) had a more complex double swing hanger system. I remember looking at the drawings of the centre truck arrangement of the passenger units in the Clyde offices at Eagle Farm and the three of us were just amazed at the complexity compared to the EMD designs we were all used to. 

Sadly, all the passenger units were either converted to freight operation (20) or scrapped (the remaining ten). The electrification remains but for most of its length it suplies only two tilting electric multiple unit trains making three trips in each direction. All freight is now diesel.

The remaining Clyde ASEA Walkers locomotives are largely still in service on coal trains.

Peter

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Saturday, April 18, 2020 11:30 AM

Duplicate post

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Saturday, April 18, 2020 11:02 AM

I was the noise engineer at the time the GM10B was designed so I did do noise testing on it but that was my only part of that project. The trucks and all the electrical gear including traction motors was an ASEA design. After the build and testing of the GM6C, which used modified HT-C trucks to accommodate the E88 motors with roller suspension bearings, EMD decided to do the GM10B for which the higher power required much bigger than the standard EMD motors. The ASEA motors used on the GM10B, and similar motors on the AEM-7 required 51" wheels so a whole new truck design was required. If EMD had designed the trucks, it no doubt in my mind would have been two 3-axle trucks but ASEA had the 2-axle truck design for their big motors so it was built in a B-B-B arrangement.

Dave

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Friday, April 17, 2020 9:38 PM

SD60MAC9500
 

Bogie_Engineer It's funny you mentioned the truck used on the AEM-7 and GM10B.. The GM10B has always been a very interesting locomotive. 10,000HP on a Bo-Bo-Bo setup. Anything of interest from the type of truck used on it? Why the choice of a Bo-Bo-Bo layout?

 

I think the GM10B used an existing design from ASEA. I have an ASEA brochure which outlined a locomotive for the Kiruna-Narvik iron ore haul which was a Bo-Bo-Bo version of the Rc4, thus about 10 000 hp. This was not built, but six Rm class were built, basically low geared Rc4s, which gave the same power in three units rather than two.

The Rms didn't live up to expectation and were transferred to the main system for freight work.

Later six axle AC locomotives using the FLEXX truck mentioned earlier were obtained to replace the existing units which were triple sets of rod coupled 1'D+D+D'1 class Dm.

Scaled down versions of the Bo'Bo'Bo' were built for 3'6" gauge in Queensland for coal traffic, but these were only 2900kW due to the small narrow gauge motors.

Peter

  • Member since
    September 2002
  • From: Sterling Heights, Michigan
  • 1,691 posts
Posted by SD60MAC9500 on Friday, April 17, 2020 8:15 PM
 

Bogie_Engineer It's funny you mentioned the truck used on the AEM-7 and GM10B.. The GM10B has always been a very interesting locomotive. 10,000HP on a Bo-Bo-Bo setup. Anything of interest from the type of truck used on it? Why the choice of a Bo-Bo-Bo layout?

 
 
Rahhhhhhhhh!!!!
  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Friday, April 17, 2020 6:47 PM

Overmod

Also -- what's your opinion of the design approach and details in the Alco Hi-Ad trucks, including what might have been done to address their issues with harmonic rock on that kind of severe cross-level profile?

 

I can't say I know the history on that truck so can't really offer any comment.

Dave

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Friday, April 17, 2020 6:45 PM

I'm not familiar with "Blomberg-P" designation but assume that is a railfan name for the GP Inclined Rubber suspension truck. That truck was fitted to some Amtrak F40PH's and some of the GP40X's on the Southern IIRC. This was the first attempt to improve the vertical ride of the GP Single Shoe truck introduced with the Dash 2 series. That truck substituted the long used elliptic springs with a pair of rubber compression pads which were shorter than the elliptics, about 5" versus 9" and those allowed the use of a shorter swinghanger, opening up sufficient clearance under the spring plank to the rail to run a slack adjuster connecting the brake levers on each axles 1 and 2. The original design suspension with leaf springs by Mr. Blomberg had 4" of secondary static deflection and 3.25" of primary static deflection. The rubber spring on the new in 1972 compression rubber suspension had only a little more than 1" of static deflection, bringing the total to about 5.25". This absolutely ruined the good vertical ride of the truck, which RR's started complaining about almost immediately.  Effort was made to try to change the damping of the new vertical primary dampers that were added to replace the lost damping in the elliptics to improve the ride but it didn't really help. So a design change to the compression rubber was developed. It used a pair of rubber compression/shear pads placed in a chevron arrangement on each side of an new design spring plank. The deflection in the secondary increased to a bit over 2" so the ride was better but not great. This used the same bolster and the long swinghangers but required a new spring plank (the compression rubber used the original bolster and spring plank with some modifications). There were failures of the steel shims embedded in the rubber springs fixed by some mods to the plates and shimming was required as the rubber took a set. At the time, really throughout Dash 2 production, most of the GP orders were Locomotive Rebuild Orders (LRO) which reused as much of the trade-in loco as possible so EMD was doing a robust business rebuilding swinghanger trucks. Since a new, relatively expensive spring plank and springs were required, it wasn't popular with our Sales Dept. Ultimately, the supplier of the elliptic springs was able to make a lower height elliptic spring with fewer but much thicker leafs that had about 2.5" static deflection. This retained the original spring plank for LRO orders although it used the shorter swinghanger. The Low Profile Elliptic springs as they are known at EMD became the new standard GP single shoe truck suspension around 1984 and it gives a reasonably good ride, but still not as good as the original. EMD has sold a lot of the low profile springs as an upgrade since it is relatively easy to replace the compression rubber springs.

The HT-B experience wasn't a great one either. The design was started around 1974 (before my time in the Truck Group) as a way to improve the weight shift performance on the GP line. The standard GP truck has about 84% of it's static axle load on the leading axle at 25% adhesion which is pretty poor. Dofasco had their Zero Weight Transfer truck which I am sure inspired the EMD engineers. With the ZWT truck and the HT-B truck, within each truck there are nearly equal axle loads at all adhesion levels although the lead truck is at about 96% of it's static load at the 25% rating point we used. In both trucks inclined rubber springs are used to focus at the top of rail, in the ZWT they are inboard, in the HT-B they are spread out wider than the journal centers (79" on EMD trucks) on the arms of a cast bolster that pivots on a center bearing that is extended down from the bottom plate to about 24" above the rail. The rubber springs are in the form of chevrons to give the desired combination of lateral, compression, and shear stiffness. The HT-B was first and last used on 10 GP40X's for UP and SP, Santa Fe got standard GP single shoe trucks and the Southern the inclined rubber. The HT-B worked as intended re weight shift but it had some rubber spring failures and lateral ride complaints that could have eventually been solved I'm sure. What killed it was a bunch of factors:

- With Super Series wheelslip control introduced on the GP40X, even though the HT-B weight shift was far superior to the GP truck, head to head adhesion testing showed it offer no benefit in tractive effort; it was theorized the lighter lead axle slipped more but cleaned the rail more effectively for the trailing axles.

- It cost a lot more to produce - with any new design, the suppliers have the ability to charge for the newness compared to the very mature cost of the design in production for 40 years. The factory didn't like fabricating the underframe with the dropped center bearing. The truck assembly was also heavier by several thousand pounds, not a good thing for 4-axle locomotives.

- The biggest drawback was it killed a major cost saving on an LRO order where the GP truck was sent for scrap instead of being reused.

So even though it showed up in GP50 literature as an available option, no RR wanted it and EMD had no desire to produce it after the GP40X's. There was one spare frame in production inventory in 1984 when I bought it to build the first radial test truck.

Dave

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Friday, April 17, 2020 2:07 PM

What is the specific history (and experience) of the use of 'chevron springing' on the GP40X HT-B and then laterally on the "Blomberg-P" trucks?

Also -- what's your opinion of the design approach and details in the Alco Hi-Ad trucks, including what might have been done to address their issues with harmonic rock on that kind of severe cross-level profile?

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Friday, April 17, 2020 12:38 PM

The long travel secondary springs provide a great vertical ride and low bogie yaw stiffness, but in a 3-axle truck, the short travel primary springs would not work in North America where staggered joints are prevalent, they just can't equalize well enough. Our design criteria for any bogie was to negotiate a 39' staggered rail joint profile with a 3" inverted sine profile. After center axle spring pocket cracking on SD Flexicoil trucks, on Rock Island I believe, the criteria was raised to 4" for switcher applications. Stiff primary just won't make it.

We used to refer to our associate Henschel's Flexi-Float bogie as having a wagon tongue. For higher speeds, the rubber chevron springs do a great job of locating the axle and give high axle lateral stiffness necessary for stability. Our only EMD experience was on the AEM-7 and GM10B ASEA designed trucks. They'd never work here in a 3-axle truck due their short travel, about 1.25". We did try to use them on the HT-B truck but they had a lot of set and drift issues there.

Dave

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Friday, April 17, 2020 2:13 AM

Overmod

Neatly in the middle between the HTCR and UGL trucks was the Adtranz/Bombardier "FLEXX", as mounted on  the Blue Tiger locomotive; this as I recall it was a kind of antithesis of soft primary/stiff secondary (with nested harmonic-breaking primary springing, very long secondary springs, and a traction strut as low physically in the locomotive as possible).  There was at one time an interesting technical paper (as I recall, for one of the conferences) that went into design detail, including as I recall the line of development from the Henschel Flexi-Float bogie design, and someone astute may still be able to pull this out of the Wayback Machine.

I can still remember when the Fabreeka-style chevron springs were supposed to be the future of controlled primary suspension...

 

The Blue Tiger was a bit of a flash in the pan...

I have a nice HO model of the prototype....

But here is a brochure with a good drawing.

https://web.archive.org/web/20081121082622/http://www.bombardier.com/files/en/supporting_docs/BT-Bogies-FLEXX_Power.pdf

The big flexicoils were very popular as secondary suspension for years.

Peter

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Thursday, April 16, 2020 11:24 PM

Neatly in the middle between the HTCR and UGL trucks was the Adtranz/Bombardier "FLEXX", as mounted on  the Blue Tiger locomotive; this as I recall it was a kind of antithesis of soft primary/stiff secondary (with nested harmonic-breaking primary springing, very long secondary springs, and a traction strut as low physically in the locomotive as possible).  There was at one time an interesting technical paper (as I recall, for one of the conferences) that went into design detail, including as I recall the line of development from the Henschel Flexi-Float bogie design, and someone astute may still be able to pull this out of the Wayback Machine.

I can still remember when the Fabreeka-style chevron springs were supposed to be the future of controlled primary suspension...

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Thursday, April 16, 2020 9:59 PM

Interesting paper Peter. The basic design of the UGL bogie is really no different than the latest EDI and EMD GFC bogie where all have adopted axle guidance via fixed traction rods and soft primary, stiff secondary vertical suspensions. The authors dismiss the radial steering bogies using linkages which serves their purpose but the real world improvement in wheel wear and lateral rail force is well documented in technical papers by EMD and AAR. While it is true that the steering of the axles declines with higher tractive effort, the flange force also declines so flange wear is improved. It's worth noting that locomotives don't spend the majority of their operational time at the limits of adhesion where the friction is saturated. As the paper states, it is true that customers that previously bought GE's Steerable Truck on their locos have gone back to the standard "roller blade" trucks, however, EMD's customers continue to purchase HTCR radial trucks even with a cheaper non-steering version available.

Dave

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Thursday, April 16, 2020 8:51 PM

Dave,

Thanks for the response.

As well as the EMD associate, the GE associate UGL have developed a similar bogie to the last EDI design.

railknowledgebank.com/Presto/content/Detail.aspx?ctID=MTk4MTRjNDUtNWQ0My00OTBmLTllYWUtZWFjM2U2OTE0ZDY3&rID=MjQ3Mg==&sID=NQ==&ph=VHJ1ZQ==&qcf=MTk4MTRjNDUtNWQ0My00OTBmLTllYWUtZWFjM2U2OTE0ZDY3&bckToL=VHJ1ZQ==

This is called the "Flexicurve".

The paper has some diagrams that might help explain to other readers the considerations in the designs being discussed here.

UGL have not applied this design in Australia yet, all the locomotives (including the C44aci, the competitor to the GT46C-ACe) using a familiar GE export design.

UGL have supplied many trucks for GE export locomotives all over the world, and they must be regarded as successful if not reflecting recent ideas.

Peter

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Thursday, April 16, 2020 6:58 PM

M636C

I've found a presentation with an excellent photo of the GT46C-ACe truck...

https://web.archive.org/web/20090913111026/http://www.core2008.org/assets/papers/tuesday/1045/Paul-Hewison/Presentation%20-%20DEDIR%20SCT%20HALCROW%20-%2030-8-08%20v2.pdf

This confirms that the truck does not use a steering mechanism...

Page 18 of the presentation.

Peter

 

Thanks for the link, had not seen that presentation. 

You are right that this truck does not use a steering linkage but it is considered self-steering none the less. Here they use a swingarm-type design with a special design bushing at the pivot that allows some longitudinal travel for steering. With this arrangement, under high tractive effort, there is longitudinal translation of the end axles on tangent track but the axle will steer itself in curves. What we found during testing, where we instrumented the axle traction rods as force transducers, was that if we constrained the steering beams to not allow axle steering, the forces in the traction rods during coasting conditions were nearly equal and opposite. The taper on the wheels is trying to steer the axle into a radial position with the outer traction rod in the curve being in tension and the inner in compression. As tractive effort increases, the tractive force adds to the steering forces in the traction rods such that the inner rod may see a very low force and the outer one very high force. With a pedestal truck, no motion is allowed so the pedestals have greatly different forces acting on them in curves which is not apparent.

Back to the EDI design; so with the relatively soft longitudinal bushings, the axle can steer itself within the limits of travel allowed by the bushing design. A truck of this arrangement may not fully radial steer in 10 degree curves as the steering beam arrangement allows, but it will be able to steer radially in up to 6 or 7 degree curves, which is still beneficial.

This same arrangement in a little different form is used on the EMD GFC fabricated frame bogie used on export locomotives such as GT38AC thru GT46AC. That bogie is available with soft and stiffer bushings depending on the service and customer needs.

In the first EDI arrangement with the cross-beam for the carbody connection, an advantage of the that design is that it transmits the tractive effort into the sidesill where that underframe has enough strength, unlike the traditional EMD underframe design where the strength is mostly in the centersills.

You'll note that none of these newer designs have an inter-connection between the end axles linking steering motion. I'll discuss this in the last installment in this thread.

Dave

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Thursday, April 16, 2020 8:17 AM

I've found a presentation with an excellent photo of the GT46C-ACe truck...

https://web.archive.org/web/20090913111026/http://www.core2008.org/assets/papers/tuesday/1045/Paul-Hewison/Presentation%20-%20DEDIR%20SCT%20HALCROW%20-%2030-8-08%20v2.pdf

This confirms that the truck does not use a steering mechanism...

Page 18 of the presentation.

Peter

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Thursday, April 16, 2020 12:37 AM

Dave,

Thanks for the post and particularly the link to the patent and the drawings.

One of the differences from the Australian trucks was the flexible link to the centre pivot which was quite different.

In the Clyde/Downer design, the truck centre pin was located by the external brackets and links to a transverse beam seen in the photo of the narrow gauge version on S 2106 illustrated here. The standard gauge locomotives of classes Q, FQ and V had generally similar arrangement, as did the fairly successful narrow gauge GT42CU-AC of which 250 or so have been built, the most successful cape gauge locomotive ever built in Australia.

All these trucks used the four rubber/metal secondary suspension pads as shown on the EMD patent. I think the MLW/ Dofasco truck was the first to use these pads in 1967, but now both EMD and GE use these pads.

Downer changed to the EMD design of centre pivot link as shown in the patent with the GT46C-ACe which was basically an SD70 ACe squeezed into the Australian clearance gauge.

While not as clear as the S class photo, it can be seen that the external brackets and links are not present on the TT.

Downer indicated that the GT46C-ACe trucks were not strictly steering trucks since there was no steering linkage, but since similar bearing and motor mountings were used, the axles were free to take up radial locations in curves.

I was fortunate to be in Melbourne when the first GT46C-ACe units made their first run in service.

I turned up along with quite a crowd, including engineers from EMD, who I asked about the new trucks. I was quite surprised to be told that no instrumented tests had been performed, but they had run high speed test runs that indicated no hunting or other operational problems. (Among the crowd was the CEO of the operator, Mr Smith of SCT, who seeemed to be having a good time, but he did not ride the train, returning (presumably) to the office by car).

Peter

  • Member since
    April 2007
  • From: Bridgman, MI
  • 283 posts
Posted by bogie_engineer on Wednesday, April 15, 2020 8:14 PM

Thanks for the positive comments.

Continuing the radial truck development story...

When we got back from the successful 2-axle radial truck test, we immediately started to brainstorm how we could do a 3-axle truck. After much discussion, we settled on an arrangement that retained the steering beams for axle location and yaw freedom, along with an inter-connecting link running across the top of the middle traction motor. The rubber bumpers limiting steering beam rotation and thus axle yaw were retained. We looked at using bellcranks, as GE eventually did on their steerable truck to avoid EMD's patents, but rejected them due to the difficulty cross-linking them and the greater number of pivot bearings adding cost.

Although it would have been simpler to use a motor arrangement like the SD Flexicoil truck where the outer motors face inward, we kept the HT-C arrangement for its superior weight shift - at 25% adhesion, the lightest axle on the SD Flexicoil truck is at only 78% of its static load, whereas, on the HT-C and now the HTCR, it's at 93%. We knew we needed to retain the feature of truck yaw stiffness to keep the truck tracking straight on tangent, yet needed the high compression stiffness of the secondary springs to keep the truck from pitching to get the minimum weight shift. That led to the tall rubber springs that shear about 8 inches at full truck rotation, not unlike the Dofasco Hi-Ad truck. Round springs gave us a relatively soft lateral stifffness for good ride and soft rotational stiffness necessary for good curving. With this direct support between truck frame and underframe, no bolster is needed, however, a means for transfering the tractive effort of the truck to the underframe is required. This was accomplished with a four bar link arrangement connecting the first transom thru traction rods to a pivot block that rotates about a post extending down from the underframe and is detailed in the patent referenced below. Elimination of the bolster opened the space for the inter-axle link to run above the middle traction motor.

We replaced the rubber traction motor nose support pack, used since the early days at EMD, in favor of a link with rubber bushed ends. This was done for two reasons; first, to get rid of all the wearing parts that including welded wearplates and pins, and secondly, to have a consistent stiffness of axle rotation unaffected by unpredictable friction in the nose pack.

The primary suspension used long travel single coil springs for good vertical ride and axle load equalization on a new bearing adapter casting. A single rather than double coil was used to avoid coil binding during axle yaw and was made possible by spacing the springs away from the axle centerline allowing for taller springs. Additional primary suspension dampers were added to make up for the loss of friction damping provided by the pedestals in conventional trucks. To react the lateral force of the axle against the truck frame, a steel faced rubber pad mounted on top of the bearing housing reacted against a nylon wearplate mounted inside the truck frame between the coil springs. Integrally cast secondary lateral stops on top of the center axle spring pocket react against a wearplate welded to edge of the bottom plate. Blocks welded to the bottom plate limited truck rotation to 6.5 degrees.

Brake rigging was similar to the HT-C truck but we put a brake cylinder adjacent to each wheel which allows enough piston travel to fully wear out a brake shoe so that no adjustment of the brake rigging is needed between shoe changes, a perceived maintenance advantage.

The patent for the HTCR is here showing the truck arrangement: http://www.freepatentsonline.com/4765250.pdf

In about six months from the finish of the 2-axle test, we had the design far enough along to begin procuring parts to build the first HTCR truck. The major part was a new cast truck frame - working with our long time supplier, the former LFM in Atchison, KS (at the time part of Rockwell International), we designed a solid steel cast frame that had no internal coring. This added significant weight, almost 9,000 lbs. more the HT-C bare frame, but it saved us substantially in pattern cost. They built a simple pine pattern and a few coreboxes for some frame external brackets for $125K, a full hollow frame pattern would have cost at least $300K and since it hadn't been tested, we weren't completely confident this would be the final design. Rockwell did a great job and we had a casting to build the truck in about 8 months, a very short time for this kind of procurement.

New axle bearing housings were cast at Racine Steel Castings and are the only other cast parts in the truck. Lord Corporation in Erie, PA made most of the rubber parts and Hadady Corp., EMD's brake rigging supplier made the steering beams, brake rigging, and numerous other fabricated and machined parts. 

In 18 months, we had the test truck assembled and on it's way to Pueblo for testing. The EMD Experimental test shop fabricated an adapter structure that was necessary to mount the truck under SD60 EMD3 that we used for testing. This structure incorporated a traction post welded into a 3" plate that connected the truck to the underframe. The truck assembly and the adapter are shown in the only two photos I have from this phase of the project here: https://flic.kr/s/aHsmMz7Mt3

When we had all the material at TTC in fall of 1987, we placed the adapter on top of the truck assembly, raised the locomotive on jacks, rolled the truck under the unit, and lowered the unit onto the truck. Once positioned where we wanted it, the adapter plate was welded to the bottom plate of EMD3. 

The test program began at TTC for preliminary curving tests up to 7.5 deg curves, the max at the time at TTC, dynamic response on the perturbed tracks, and hunting stability testing. The dynamic tests of Dynamic Curving, Twist and Roll, Yaw and Sway, and Bounce showed no problems. Stability speed, however, did not meet our goal - at about 55 mph we experienced significant truck hunting involving both truck yaw and axle yawing. We experimented with different axle yaw damper characteristics but did not solve this issue during this test program, that would have to be studied more and figured out after this test was completed.

We moved on to Raton Pass for the curving portion of the test. For testing, we built a feature into the truck so we could lock the steering to get a good A-B comparison of rigid versus radial steering. We also moved the instrumented wheelsets to the HT-C truck so we had a direct comparison to that truck. Key test parameters were Net Lateral Load (NLL) at the leading outer wheel in the curve, Angle of Attack (AoA) between the leading outer wheel and the rail, and a flange wear index, calculated by multiplying the lead outer wheel flange force by the AoA. We made repeated runs up the hill from Trinidad to Raton tunnel stopping short of the portal after the big sweeping right hand curve at Wooten Ranch.

The test results exceeded our expectations; after averaging many test runs, the data showed for 10 degree curves that the NNL of the radial truck was reduced about 30% coasting and 20% at 30% adhesion compared to the HT-C truck. Angle of Attack was reduced about 90% coasting but only about 10% at 30% adhesion. The flange wear index remained nearly constant for the HTCR truck over the range of adhesion levels as the AoA increased with adhesion level that is nearly offset by a lowering of flange force as more of the available friction is used for traction and less is available to steer the wheelsets. For the HT-C truck, the AoA remained constant with adhesion level but the flange force declined resulting in a about a 30% reduction of flange wear index at 30% adhesion. The end result was that the flange wear index for the HTCR truck was reduced by about 60% coasting and 30% at 30% adhesion compared to the HT-C. These results were presented in detail in a technical paper published and presented at the 4th International Heavy Haul Conference in Brisbane, Australia in 1989.

So at the conclusion of the test in early 1988, we had a 3-axle radial truck that steered very well but needed work on hunting stability before it was ready for production.

Dave Goding

 

  • Member since
    September 2002
  • From: Sterling Heights, Michigan
  • 1,691 posts
Posted by SD60MAC9500 on Wednesday, April 8, 2020 8:52 PM
 

 

Great thread Bogie_Engineer! It's always a pleasure to get this info from those directly involved. I just learned more in your thread about dynamics than all of my high school teachers who taught this very subject!

 
Rahhhhhhhhh!!!!
  • Member since
    December 2017
  • From: I've been everywhere, man
  • 4,269 posts
Posted by SD70Dude on Wednesday, April 8, 2020 6:44 PM

Thanks for that.  Wow, seven computers.  What a kludge. 

Building modern passenger locomotives sounds about as easy as operating passenger trains, the most effort for the least profit. 

Now back to your regularly scheduled thread, this stuff about radial trucks is fascinating.  I'll put on some more popcorn and leave you guys to it.

Greetings from Alberta

-an Articulate Malcontent

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy