External and Internal Features Added To UP Locomotives
Index For This Page
This page was last updated on January 24, 2016.
The first GP7s, numbered UP 700-709, delivered in February and March 1953, had their horns mounted near the rear exhaust stack (long hood forward).
Photos of UP's fleet of GP9s when they were new show that their air horns were originally located at the top of the cab roof, at the centerline of the locomotive.
In the mid 1960s, UP began changing the horn location from the cab roof, to above the radiator fans. The new location was meant to alleviate the problem of horns not operating properly due to buildup of ice and snow. In a series of engineering drawings dated May 1963, the 1-30 series GTE locomotives, the GP7s and GP9s, the GP20s, the GP30s, and the SD24s, and the Alco DL-640s, were all the subject of drawings showing new air horn locations to prevent freezing. (ST 8086, 8088, 8091, 8099, 8105, 8106, 8108)
The first new units with the new horn location were the 22 GP35s, UP 740-761, delivered in May and June 1964. However, the GE U50s, also delivered in 1964, came with cab-mounted air horns. New DDA35 units delivered in 1965 had rear-mounted horns, as did the SDP35 units also delivered in 1965.
All subsequent locomotive deliveries after mid 1964, and until early 1978, had their air horns at the rear, mounted above the radiator fans. Locomotives built before 1964 were modified to match the new standard by installing a new horn bracket and running an air line (usually 1/2-inch copper) from the former location, along the roof to the new location.
Beginning in 1978, new units were delivered with their horns mounted on the cab roof. The location above the radiators became a problem after studies in 1977 found that the rear mounting location reduced the horns' effectiveness, with the locomotive body and obstructions along the roof deflecting and masking some of the sound waves. In-service units received the change over the following years, as scheduled maintenance took place.
For new locomotives delivered in the 1972-1974 time period (especially the SD40-2s and GP38-2s), the air supply line was installed along the locomotive frame from the area under the cab floor, back to the rear of the locomotive, where a vertical pipe was installed to the roof, adjacent to the sand fill, with a supply line mounted to reach the horn location between the radiator fans (this can be seen in photos).
After the 1978 official change of horn location, the air supply line was extended along the roof on the right side of the radiator fans and dynamic braking fans, to the new location on the cab roof. Relocated air horns were mounted to a raised bracket on the cab roof. Units delivered after the 1978 change had their air horns mounted lower and closer to the surface of the cab roof.
Moving the air horns from the rear of the units, to the cab roof by installing an extended air line along the roof was soon found to be both redundant and expensive. The modification program was changed to retire-in-place the existing air horn supply line, and install a new line from under the cab floor, up through the electrical cabinet to the cab roof. This is seen most notably on SD40-2s and GP38-2s, on units with cab-mounted air horns, along with an unused bracket between the radiator fans, and the short air supply line from the unused bracket back to the rear of the unit's roof. Photos show a wide variety of locomotives of these two types, with variations of the rooftop external air lines. As a general statement, SD40-2s in the 3100, 3200 and 3300 series, and the 60 GP38-2s in the 2000-2059 series, were delivered with rear mounted air horns, and later received external rooftop air lines, although many also show the retired-in-place bracket and short supply line only at the rear. SD40-2s numbered higher than the 3420-series were delivered with cab mounted air horns, which can easily be seen by the fact that their air horns are mounted low to the cab roof.
Photo research suggests that the changeover for new units came in March 1978, with the delivery of SD40-2s 3410-3488. A photo of three new SD40-2 units on page 160 of George Cockle's Union Pacific 1977-1980, shows UP 3414 and 3415 with their horns back among the radiator fans, and UP 3416 with the horn in the new location on the cab roof. For approximately 15 units, the change order came too late for EMD to re-design the internal piping that furnished air to the rear location. As an interim fix for approximately 15 units delivered in March and April 1978, EMD put a large air supply line along the entire length of the roof, from the rear location to the new location on the cab roof. (UP's six GP40Xs, numbered as UP 9000-9005, were delivered in March 1978, with their air horns mounted above the radiators.)
UP's first order of 15 C30-7s, 2960-2974, were delivered between July and September 1977. As delivered, these first C30-7s had their horn located on the left side of the forward section of the radiator housing. Within a year the horns on other GE units were relocated to a position just behind the operator cab to maximize the forward projection of sound. The second C30-7 order, UP 2415-2429, were delivered in June and July 1978, with their horns mounted on the roof immediately behind the cab.
"UP used to have the Leslie S-3L on everything, but in recent years they have been buying the Nathan (AirChime) K-3HA. This is what the SD9043MAC's were delivered with. Last year, I believe UP announced they were switching to the K-3HA exclusively due to complaints about how lousy their horns sounded (the Leslies don't stand up well to the abuse of button-controlled horn valves)." (Posted by Evan Werkema on Trainorders.com, February 2, 2000)
UP 3049 was equipped with an electronic siren. The following comes from UP Eastern District Bulletin A-300, dated August 6, 1975:
Unit 3049 has been equipped with an electronic siren on a test basis. This siren is to be used as an emergency vehicle only and in no way alters the requirements on use of the regular horn. The operator has a selection of three sounds: WAIL, YELP or HIGH-LOW and may be used individually or in a series. It is intended that the siren be sounded approaching and passing gangs or workmen on or near the track and is not intended to be used in cities, towns , municipalities or grade crossings except in emergency. (courtesy of Mark Amfahr, email dated September 17, 2008)
UP SD40 3049 was involved in test program to use air raid-type for sirens to better alert track crews. (CTC Board, January 1976, page 8)
UP SDP35 1402 was equipped with air raid-type siren, made by Federal Signal, to increase alertness of track maintenance crews. SDP35 1400 also equipped with same design. (Pacific News, May 1978, page 16; Pacific News, August 1978, page 15)
As part of the same tests to improve warning devices for track crews, UP SDP35 1408 was equipped with two air horns mounted on its cab roof. (see photo of UP 1408, on June 2, 1978, on page 114 of George Cockle's book Union Pacific 1977-1980, published in 1980)
Siren was removed from UP SDP35 1402. (Pacific News, January 1979, page 14)
Air raid type sirens have been removed from SDP35s 1400 and 1402. They will be reinstalled on 6900s. The first to get them was 6918. (Pacific News, March 1979, page 16) UP 6918 retained its siren while in storage in Yermo, California, from January 1982 through its retirement in 1985.
Air raid siren from SDP35 1400 has been installed on 6924. (CTC Board, May 1979, page 8; Pacific News, May 1979, page 22)
Automatic Cab Signals
A note on terminology:
- ACS denotes Automatic Cab Signals -- located trackside, on the ground; noted in employee timetables.
- CCS denotes Coded Cab Signals (full name: Code Cab Signals-Safety Control, CCS-SC) -- the equipment located on the locomotives
Union Pacific's ACS/CCS system used a design from Union Switch & Signal which first became available in the mid 1920s, and which was first installed in its earliest forms on Pennsylvania Railroad in 1923 and 1926. ACS/CCS uses a four-light cab signal and locomotive-mounted induction receivers that sense a pulse code that is fed into the track itself.
In 1949, Union Pacific first added Automatic Cab Signals as part of a general improvement of traffic control along the railroad's Oregon-Washington Railroad & Navigation Co., mainline between The Dalles and East Portland, via Graham (85.3 miles). The system was used in conjunction with Automatic Block Signals to improve the flow and control of traffic by installing duplicating signal indicators inside the locomotive cabs, with an acknowledging action required from the engineer upon receiving an more restrictive signal indication. The greatest advantage of the ACS system was that it allowed trains to operate at maximum speed in inclement weather, without concern of visibility of trackside signals. The line was called the Oregon Fourth Subdivision.
UP has ACS on the Portland Sub between Crates, the west end of double track at The Dalles, through to Troutdale. The previous edition of the timetable also showed ACS on the Graham line from Troutdale to East Portland but the current edition doesn't show it, just CTC. (Bill Maltby, March 2000)
In January 1950, UP's engineering department issued a drawing showing the installation of four-indication light boxes in the cabs of the Alco freight units, running on the former LA&SL west of Salt Lake City. In February and April, a drawing was issued covering all of the railroad's diesel passenger units, showing the wiring changes needed to use four-indication light boxes on locomotives with three-indication light systems.
ACS was installed along the Nebraska and Wyoming mainlines, west into Ogden, Utah, after a terrible accident at Wyuta, Wyoming, on November 12, 1951, where one of UP's Streamliner passenger trains ran into the rear of another Streamliner train. Train 104, the City of San Francisco traveling at 77mph, rear-ended Train 102, the City of Los Angeles. Train 102 had just started to move again after stopping at a signal light that had become obscured by snow and ice. (see also: Railway Age, Volume 131, November 19, 1951, page 13)
The following comes from the May 1972 issue of Railroad magazine, page 59:
How Cab Signals Work
UNION PACIFIC acquired, in February, 70 sets of type EL coded cab signal equipment from Westinghouse Air Brake Co. for installation on new locos. This consists of a small signal unit mounted at the engineer's eye level, plus sensor equipment ahead of the front wheels. It provides a continuous indication of track conditions ahead. To inform readers interested in modern cab signaling, this explanation comes from R. J. Casey of WABCO:
Continuous cab signals are operated by processed data received from the coded alternating current transmitted down the rails towards the train. By means of receivers mounted ahead of the leading wheels, the coded signal is picked up and converted into a visual signal on a small display panel in the driver's cab.
Codes are obtained by varying the pulses of alternating current by means of relay contacts and are measured in terms of the number of pulses per minute. A clear wayside signal indication is equivalent to 270 pulses per minute and shows a green light on the cab signal. Three degrees of speed reduction, signaled by illuminated signs on the display panel, are triggered by 180, 120, or 75 pulses per minute, depending on the number of clear track sections ahead. A stop indication is ordered when no pulses are received.
When the change is to a more restrictive signal, an audible warning signal sounds in the locomotive. Fail-safe reliability is ensured by the inherent design of equipment and circuits. Any malfunction automatically produces the most restrictive indication. Engine carried apparatus responds uniquely to pre-established code frequencies- providing immunity from foreign current.
Voltages of the same frequency and code rate, are induced into the receiver coils on the locomotive and after amplification, are used to operate a master relay. This relay governs four decoding units that are selectively operated. The associated decoding relays in turn control the cab signals and the warning bell.
Intermittent cab signals are actuated through a pair of coils placed along the track rather than by the transmission of current through the rails. Various aspects of wayside signals are connected to condensers that tune the coils to the required number of frequencies. Frequency emissions are detected by locomotive mounted receiver coils as the train passes over trackside coils.
Trackside installation consists of a high and low frequency coil while a corresponding pair of wide band amplifiers are mounted on the locomotive. Amplifier oscillations are measured by low and high frequency detectors that pass the corresponding steady signals to the logical circuitry. Circuitry is equipped with fail-safe solid state components that compute the incoming signals and trigger the various signals within the cab and, in certain circumstances, set off operation of train control equipment. Use of high and low frequency coils permit transmission of a broader combination of track condition information.
As of March 2000, UP had ACS in the following locations:
- Summit (Omaha, Nebraska) to Riverdale CPC 988 (near Ogden, Utah)
- Laramie to Green River completed before September 1951
- Cheyenne to Laramie, and Green River to Ogden added by April 1961
- Menoken, Kansas (near Topeka) to Gibbon, Nebraska
- O'Fallons, Nebraska (just west of North Platte) to near South Morrill, Nebraska
- The Dalles, Oregon, to East Portland (Troutdale), via the Graham line, added in about 1949 (earlier versions install as early as 1936)
- CCS equipment was added to C&NW locomotives to allow them to lead UP pool trains between Fremont and North Platte, Nebraska.
- (Mark Amfaur, email dated March 28, 2000)
Automatic Train Stop (AT&SF)
"UP E units were equipped with Coded Cab Signals (CCS) for the Overland Route main line. The pickup bar was located under the pilots. The ATSF between Daggett and Riverside was equipped with a different system, Automatic Train Stop (ATS), in order to comply with FRA requirements to exceed 79 mph running between Victorville and Barstow. Some UP E8's and E9's appear to have a Santa Fe style ATS pickup shoe on the right side of their 5th axle." (Olin Dirks, email dated September 17,2009)
According to Wikipedia, the federal Interstate Commerce Commission required ATS after 1951 as a minimum safety requirement to allow passenger trains to exceed a speed limit of 79 mph. The ICC decision followed the terrible wreck at Naperville, Illinois, on April 25, 1946, in which a CB&Q passenger train traveling at 85 mph ran into another CB&Q passenger train that had stopped due to a mechanical problem. The regulatory requirement refers to a system that triggers an alert in the cab of the locomotive whenever the train passes a restrictive wayside signal and that then requires the locomotive engineer to respond to the alert within a set period of time before the brakes are automatically applied. The most popular implementation of ATS for the mainline railroad industry was made by the General Railway Signal company starting in the 1920s and consisted of inductive coils mounted just outside the right hand rail in relation to the direction of travel. Often referred to as just ATS in railroad operating books, the full name of the system is Intermittent Inductive Automatic Train Stop to differentiate it from mechanical systems being offered at the time. AT&SF installed ATS along its Chiacgo to Los Angeles route at points where train speeds exceeded the 79 mph requirement.
According to an internal company roster dated September 1, 1968, all of UP's E9s (UP 900-914, 943-961) and all of their E8s (UP 925-942) were equipped with ATS.
I don't have anything from earlier times, but it is likely that only certain units had ATS, and by the late 1960s, all potential lead units had been modified.
Automatic Train Control (C&NW)
Coded Cab Signal (CCS) is the Union Pacific's cab signal system and is used on the UP west of Council Bluffs. It is similar to ATC, but with several differences. Wayside signals exist in CCS territory. The CCS cab indicator has four aspects (clear, advance approach, approach and restrictive) as compared to ATC's two. And CCS track signals are coded using different frequencies, whereas ATC's track signal is non-coded. Using CCS equipment from UP SD45s retired in 1981 and 1982, C&NW GP50 5063 was first equipped with CCS on top of ATC on a test basis in 1983. Dual ATC-CCS applications were later extended to all GP50s (equipment later transferred to 48 SD40-2s), all SD50s, 31 of the SD60s, and almost all of C&NW's GE locomotives. The presence of dual ATC-CCS permits C&NW locomotives to lead trains over both the C&NW and Union Pacific east-west main lines.
As part of the Amtrak Authorization and Development Act of 1992 (Public Law No: 102-533, October 27, 1992), the Federal Railroad Administration formally instituted a regulation that required ditch lights on all locomotives operating in the United States. The regulation required that by 1997, any locomotive operating faster than 20 miles per hour over public grade crossings must be equipped with a triangular-pattern of lights; one headlight and two "alerting" lights. Ditch lights had been in use for the past 40 years on Canadian railroads as an aid to crew safety, and are used to illuminate the side portions, the "ditches," of the railroad right of way on mountain lines with numerous curves. Based on statistics, the FRA hoped to eliminate as many as 3,300 grade crossing accidents over the following 20 years. (see also Trains magazine, Volume 56, Number 6, June 1996, page 20)
The following comes from the May 1996 issue of PacificRailNews, page 10:
FRA Issues Ditch Light Rule -- Back in 1992, Congress directed the Federal Railroad Administration to require "substantially enhanced locomotive visibility measures"— that is, additional lights—in order to improve grade crossing safety. After several years of study and hearings, FRA published its final rules on March 6. These rules confirm the standard the industry has adopted—a triangular system of one headlight and two ditch lights. The ditch lights must be at least 60 inches apart, or 60 inches lower than the headlight. This requirement caused some controversy because Canadian National units, which often operate in the United States have ditch lights spaced 44 inches apart. FRA sidestepped the dispute by grand-fathering engines equipped with 44-inch ditch lights before May 30, 1994. Ditch lights must be at least 36 inches above the rail, except that lights on push-pull cab cars may be 24 inches above the rail. Flashing lights are permitted but not required. Locomotives operated over highway crossings at 20 mph less, and historic locomotives built prior to 1949, are exempt. All other locomotives must have ditch lights installed by the end of 1997.
Nine UP SD40-2s were equipped with what was then referred to as "Canadian-style" ditch lights in February 1978. Their numbers were UP 3396-3399 and UP 3410-3414. They were intended for assignment in the leading-position on pool trains with Canadian Pacific through Eastport, Idaho. The modifications were completed at Salt Lake City.
In November 1989, and continuing through December 1990, UP began installing ditch lights on a total of 90 former MP SD50s and C36-7s. With authority from Work Order no. 03961, there were to be 30 units modified during 1989, and 60 units modified during 1990.
In early March 2008, new EMD SD70ACe units began to appear with ditch lights on their rear platforms. The details of this are as yet unknown.
Dynamic Braking on TR5s
The following comes from "Union Pacific Switchers and Slugs" by Don Strack, 1996:
UP 1870-1877 and 1870B-1877B (eight A-units and eight B-units) were built in September and October 1951 (between UP's SW7 and SW9 orders) and were assigned EMD order numbers 6284-A and 6284-B. According to a news item in a mid-November 1951 issue of Railway Age, UP's eight TR5s were to be initially assigned to southern California. The news item stated that six of the eight units were specifically for heavy switching in the San Bernardino yards, and the other two were for helper service on the 2.2 percent grade Cima Hill, operating out of Kelso. The use of the TR5s on these helper grades allowed UP to re-assign the 1947-built Fairbanks-Morse H20-44s to the Northwestern District.
To minimize some of the costs associated with the use of what were essentially yard switchers in helper service, in late 1952 and early 1953 the TR5 A-units were equipped with dynamic braking. The addition of this EMD-designed feature included a single 48-inch fan, and four resistance grids, one for each of A-units' four traction motors, installed in an enclosure added to the top of the carbody, just ahead of the cab, blanking out and covering the front cab windows. Because the feature was meant solely to retard the downhill descent of the "light" locomotives, no provision was made for the use of dynamic braking on the cab-less booster units. Due to the limited capacity of the dynamic braking (used only on the cab units to control the speed of their downhill descent), along with the added maintenance for the dynamic braking components themselves, the usefulness of the braking was soon outweighed the associated maintenance costs. Within a short period of time, by 1956, the dynamic braking feature was deactivated, with the components remaining installed on the units, unused.
The use of TR5s on helper service eastward out of San Bernardino was reduced considerably with the delivery of GP7s and GP9s in 1953, 1954 and 1957, but they were still regularly assigned to trains needing helpers heading up Cima Hill. The use of a single TR5 set at Caliente, Nevada for the Clover Valley grade between Minto and Islen was discontinued in late 1953 or early 1954.
The use of specifically assigned TR5 (and all other) helpers in the desert helper districts was discontinued completely in February 1959, an acknowledgement to the versatility of multiple-unit Diesel operation, which allowed the "locomotive" on the head-end to be matched perfectly to the requirements of both uphill and downhill operations. The cow-and-calf sets were re-assigned to heavy duty switching service in other locations. UP 1876/1876B and 1877/1877B were reassigned to the hump yard in Pocatello, Idaho, where they were joined Baldwin AS-616s in both flat switching jobs and serving as motive power for the hump itself.
"Union Pacific at Provo, Utah stationed a pair of TR5 cow/calf sets throughout most of the 1960's and 70's. TR5's #1874 and 1875 pulled long drags of coil steel between United States Steel's "Geneva Works" and Provo Yard, and occasionally "double headed" trains of Southern Utah iron ore to Geneva from Provo Yard." (James Belmont, Trainorders.com, March 17, 2001)
In June 1974, UP 1874 and 1875 were due for Class C engine changeout, which included removal of the engine hoods. They had to remove the dynamic brake housings as well, and since the dynamic braking had long since been disconnected, it was decided to not put the housing back on. So that meant that a new slope sheet was needed to replace the dynamic brake housing.
Dick Harding (the shop side foreman) told Orin Harkey (boiler shop boilermaker) to fabricate a new slope sheet that connected to the cab, like UP 1876 and 1877 working at Pocatello. I was working running repairs, and mentioned to Orin that the original holes were likely still in the cab wall, and since he had to make a new slope sheet, he should make it slope down instead of sloping up to match the holes Omaha added in 1952 then they added the mod. He asked Harding if it was okay. Harding's response, "Make it like it was." Orin asked what to do about the blanked-off front windows. Harding said to cut out the blanks and he would have windows added, since they were identical to all other SW7s and SW9s.
It took Orin the best part of a week to fab the new slope sheets, and make the changes to the units. Orin was an old-head steam boilermaker who had come to Salt Lake City in the late 1950s, after UP closed their Cheyenne shop, and I think he really enjoyed the challenge of matching the original design. The two TR5 cab units turned out real nice. I may have put the bug in Orin's ear, but Orin Harkey did the work, with Dick Harding's okay. Dick Harding could be contrary at times with the guys who worked for him, but not in this case. (Don Strack, posted to facebook "Rails Through The Wasatch" group, December 16, 2015, in reply to James Belmont photo of UP 1874 at Provo in 1976)
EMD Cabs on MP Units
The following MP units received EMD cabs as part of wreck damage repairs while still in service on MP, prior to merger with UP. (list from Zack Hilton, via email to LocoNotes on February 2, 2001)
|Model||First Number||Second Number|
|U30C||MP 3311||MP 2976|
|U23B||MP 2256||MP 4506|
|U23B||MP 2282||MP 4531|
|B23-7||MP 2296||MP 4607|
In June 2013, Union Pacific began receiving an entire class of modern EMD SD70 units, which UP calls SD70AH (numbered from UP 8824-8996) and modern GE ES44AC units, which UP calls C45AH (numbered as UP 8052-8257). The 'H' designation denotes "Heavy." But these are only the latest of several of Union Pacific diesel locomotives delivered with extra weight to improve traction.
(Read a roster listing of C45AH units at the Diesel Shop; scroll down to UP 8052)
Just as a point of reference for how extra weight is/was added to locomotives. While working for UP back in the late 1970s, one of their SD45s came in to the shop for wreck repairs. About half of the end sheet had to be removed and replaced due to wreck damage, revealing the interior of the frame. On UP's SD45s, the end sheets were 2-inches thick instead of the more standard 3/4 inch thick, I suspect for the added weight. While the portion of the end sheet was off, I noticed inside between the frame members there were two very large cast steel blocks, about 12 inches by 12 inches by 24 inches. These would have been added during construction, and were very well attached by several large weld beads. A quick calculation, with steel at about 500 lbs. per cubic foot, means that the two blocks weighed about 2000 lbs, or an additional 4000 lbs with similar blocks assumed to be at the other end also. (Don Strack, email to Union Pacific Yahoo discussion group, February 14, 2014)
Farr grilles are a brand name of a primary air impediment (screen) design made by Farr-Air Co. of Los Angeles. Most EMD E units and F units after 1954 had this type, as opposed to the earlier horizontal design that was fabricated. The later design was stamped, and therefore much cheaper to produce. In later years, UP removed every other slot to increase air flow, in an attempt to get their E8s and E9s to run cooler. The unique snow shields atop the winterization hatches were part of this same effort. I would imagine a quick look at Railway Age or Railway Mechanical Engineer from the 1953/1954 time period might reveal a contemporary ad.
EMD first tried what railfans call "chicken wire" screen with the FTs, F3s, and early Es, then the Farr-Air horizontal type in about 1948, then the vertical Farr-Air grille from about 1954 on.
Fuel Fill Adapters
The following comes from Keepin' On Track, A Newsletter From The Locomotive Maintenance, Planning & Technology Section of Maintenance Operations, Union Pacific Railroad, Volume 4, Number 1, October 1992:
Changeover of Fueling Equipment: Buckeye to Snyder -- Union Pacific currently uses Buckeye fueling equipment to fuel its locomotives. There are approximately 450 fuel nozzles on the railroad and 6,038 locomotive adapters (two per unit).
Defective fuel nozzles currently are being rebuilt by Emco - Wheaton, makers of the Buckeye equipment. Present performance of these rebuilds indicate that they are only lasting three to six months. Emco - Wheaton has indicated that the nozzles are worn out and should all be replaced.
The annual cost of rebuilding these nozzles is $117,000. To attain full automatic shutoff, when fueling with Buckeye equipment, both the nozzle and locomotive adapter must function properly. When either of these devices fail, fuel spillage occurs.
One of the main differences between the Buckeye and Snyder systems is that all of the automatic sensing components on the Snyder are located in the nozzle, similar to fueling an automobile. On the Buckeye system, both the nozzle and adapter contain critical components that require continual maintenance to ensure efficient fueling. Union Pacific and the Southern Pacific are the only major railroads that use the Buckeye fueling system.
A detailed economic analysis has been completed on the different options available to correct this situation. The results indicate that a replacement of the current equipment with Snyder would ensure efficient fueling operations in the long term and would eliminate the current annual rebuild expense.
In addition. there would be associated savings in terms of fuel and environmental costs. As soon as funds are available, this $1,500,000 project will be implemented. In order to be prepare for the application of the locomotive adapters, a project is being implemented to drill and tap fuel tanks. This procedure is only required on EMD units and will enable the larger diameter sensing hose to be applied. Once the complete conversion is implemented. a goal of three months has been made to complete the changeover.
Natural Gas Program
Fuel Tender Program
There were different designs of roller bearing journal boxes used by EMD used on most of their early units. In later years, and because they were interchangeable, these journal boxes were seen mixed with one brand on one axle, and another brand on another axle on the same truck.
Half of the roller bearing business went to Hyatt (a division of GM), and half to Timken (a direct competitor of Hyatt's). Some have speculated that the mix was due to wartime shortages, but the reality is that General Motors was constantly being watched by the Department of Justice to guard against antitrust violations.
The square ended journal boxes could have a speed recorder drive mounted into it. Not possible with the sloped front journal box. Inside the journal box was a thrust block to absorb the lateral shock of the lateral movement of the axle (traction motor). In the square boxes the spring plates that supported the thrust block had a hole through it for the speed recorder drive. The plates in the sloped boxes had no holes. The boxes were totally interchangeable except for the option of applying a speed recorder drive.
When applying a traction motor to the position of the speed recorder drive, the axle had to have a splined hole in the center of the axle. Not all traction motors have the splined hole in the end of the axle.
On the Union Pacific all the oil bath journal boxes were Hyatt, and all the sealed grease boxes were Timken.
Commenting on the mix of journal boxes seen on GP30s, Warren Johnson wrote to the Utah Railroading Yahoo discussion group on January 30, 2009:
The answer to your question is easy. The old larger type journal boxes are J1 and J2 journal boxes. They house the journal box roller bearings. The roller bearings are in a cage inside the journal box. The race for the rollers is on the axles. The J1 box is the one with the sloped front. The J2 box is the more square one. It was made flat on the front so if it was needed, they could attach a speed recorder angle drive. A cable ran from the journal box up to the speed recorder. These boxes were the original boxes that came out on the early F units. All of the journal boxes you showed are completely interchangeable. What ever box was handy, they used.
Multiple Unit Connections
During the early years of dieselization, multiple unit control was achieved by the use of twin receptacles, one with 12 pins and the other with 21 pins. On later units with dynamic braking, the necessary field loop connector mounted atop the left side MU boxes. From the mid 1960s on, UP converted the left side twin receptacles to a single 27 pin connector, putting a blank plate over the extra location. After the conversion, the twin receptacles (21/12-pin) were painted red, and the single receptacle (27-pin) was painted blue.
The 21/12 configuration was initially used by American Locomotive Co. Locomotives with 12/21 M.U. connections also used either 6DS or 6RL air brake schedules, which also meant that they needed matching air hose connections.
When the later 27-pin connections became available, sometimes known as AAR 27-pin, the air brake schedule was updated to 24RL.
Rich Sievers wrote on Trainorders.com on February 20, 2004:
Using the photo as a guide, on the engineer's side from the top down are the Field Loop plug, the 12-pin m.u. plug, and the 21-pin m.u. plug. On the fireman's side from the top down are the Field Loop plug and the 27-pin plug (with its cover painted blue).
UP ordered the 12-pin/21-pin arrangement in addition to the standard 27-pin plug until late 1966, so the last "straight" SD40, 3082, should've been the last new unit to have it. Thereafter, UP used just the standard 27-pin plug, and many unit that still had the 12-pin/21-pin plugs had those removed.
Field Loop refers to a type of dynamic braking control that was common at the time. Today, virtually all dynamic brake control functions are trainlined through the regular 27-pin m.u. cable.
Beginning in the mid to late 1960s, there was an on-going mod that was removing the 21-12 plugs (the red ones) and blanking off the openings. At the same time, the DB field loop plug was being deactivated, with the 27 pin plug taking over the function (this required a mod to the DB control circuits). On the Geeps, there had been an earlier program to convert one of the left side plugs from 21-12 to 27 pins. This was done to match the GP30s and GP35s/DD35s. (Don Strack, posted to Trainorders.com on February 20, 2004)
North Little Rock Snowplows
Beginning in about late 1997, UP began applying a new standard snowplow to the front of its road locomotives. This new design features an almost horizontal top portion of the snowplow and differs greatly from the previous design, which had an angled top portion. Locomotives with the new design received new snowplows at North Little Rock, and at Boise Locomotive Company.
Union Pacific's engineering department issued a drawing in May 1946 covering the apparent first installation of two-way radios on Union Pacific, on 16 locomotives that other information shows were assigned to the Omaha-Council Bluffs area.
A news item in New York Times, September 27, 1947 reads: "OMAHA, Sept. 25 (AP) -- The Union Pacific Railroad announced today a $125,000 communication improvement program that includes two-way radio in diesel switch engines in several yards, two-way radio on four freight engines and cabooses and installation of an intercommunicating system at retarder yards being built at Pocatello, Idaho. G. R. Van Eaton, superintendent of the company's telegraph department, said completion of the project is expected by the end of the year." (A similar item appeared in Railway Signaling. Volume 40, number 11, November 1947, p.727, courtesy of Mark Hemphill)
Research based on an index of Union Pacific's engineering drawings, shows that the following locomotive types received two-way radio equipment on the following dates:
- May 1946 -- Ten Alco S-2 yard switchers (ST-4431)
- June 1946 -- Six Baldwin VO-1000 yard switchers (ST-4459)
- November 1947 -- Eight EMD NW2 yard switchers (ST-4823)
- July 1948 -- Two Alco RSC-2 road switchers (ST-5206)
- June/July 1949 -- 64 EMD F3 road freight locomotives (delivered during 1948) (ST-5518; ST-5551)
- June 1950 -- 16 EMD NW2 yard switchers (ST-5905)
- November 1950 -- Two Fairbanks-Morse yard switchers, and four Alco S-2 yard switchers (ST-6028; ST-6029)
- June 1951 -- Six Baldwin AS-616 road switchers (UP 1260-1265) (ST-6328)
- September 1953 -- Five Alco S-2 yard switchers (ST-6710)
- December 1955 -- Two Fairbanks Morse passenger units (UP 654, 657) (ST-7146)
- September 1963 -- EMD yard switchers (ST-8149)
- September 1963 -- UP 1809 (EMD SW7; assigned to Camas Prairie) (ST-8150)
- December 1963 -- UP GP7 102 and F9s 503, 505, 518, 527, 539, 540 (assigned to Spokane International) (ST-8171; ST-8174)
- March 1964 -- UP 1876 and 1877 (EMD TR-5) (ST-8213)
- November 1964 -- E8 and E9 (ST-8324)
- November 1964 -- 1100 class yard switchers (Alco S-2) (ST-8329)
- December 1964 -- GP7 road switchers (except UP 102) (ST-8334)
- January 1965 -- GP20 road freight (ST-8336)
- January 1965 -- E9 passenger units (ST-8337)
- January 1965 -- 633, 637-640 (GE U25B) freight units (ST-8340)
- January 1965 -- GTE 1-30 freight units (ST-8342)
- January 1965 -- GP30 and GP35 freight units (ST-8343)
- February 1965 -- 31-45 (GE U50) freight units (ST-8348)
- February 1965 -- 1280-1295 (Alco RSC-2, RS-2) freight units (ST-8355)
- March 1965 -- 675-678 (Alco DL-640) freight units (ST-8357)
- April 1965 -- 300-349 (GP9) freight units (ST-8368)
- September 1965 -- 1211-1222 Spokane International Units (RS-1) freight units (ST-8407)
- June 1966 -- 2900-2909 (Alco C-630) freight units (ST-8497)
- July 1966 -- 2800-2804 (GE U28C) freight units (ST-8516)
- July 1966 -- 625-632, 634-636 (high nose U25B) freight units (ST-8524)
- August 1966 -- 60 and 61 (Alco C-855) freight units (ST-8531)
- (No additional drawings were apparently issued after these in 1966, showing specific reference to the installation of radio equipment; through drawing ST-9797, dated October 1978)
Union Pacific issued a separate letter with each locomotive order, listing the specifications showing how each order would be equipped from the factory, and what UP would add at Omaha during the set-up. The letter for GP9s 300-349 in April 1957 does not mention radio equipment, including antennas. Research using photographs suggests that all these early installations used the common 12-inch whip antenna
A similar letter of setup instructions for the 3600-3649 SD45s in January 1968 says that the voice radio should be installed the same as SD40s 3000-3082, delivered in March 1966. It also states that the radio antennas are "Firecracker type, Model ASP-16," without any information about where to put them.
In later letters, as late as December 1974 (the most recent available) the instructions are merely to install the radio on top of the control stand, the same as for SD40s 3000-3082. So, it appears that the SD40s delivered in March through December 1966 were the touchstone of radio installation, which might explain why the above list ends with the same year, concerning individual drawings for individual classes of units.
Radiator Fans, 48-inch Fans on GP9s
(First published to UtahRails.net blog on April 15, 2011)
For several years from the early 1970s through to the late 1990s, I thought that Union Pacific GP9 cab unit number 321 was the first GP9 on UP to receive 48-inch radiator fans. I was mistaken.
Back in 1972 or 1973, I was in the EMD parts warehouse in Ogden, Utah. In the foyer, they had a full set of EMD parts catalogs (in the standard, flip-up EMD catalog rack). In the same rack, there was a set of an EMD sales department newsletter, called "Inside EMD." It contained all kinds of interesting stuff, especially from the early 1950s. One bit of information I've always remembered was it called out specific orders when options were available, or no longer available.
On the subject of "Inside EMD", Preston Cook, a historian of EMD locomotives, wrote:
Inside EMD was originally introduced as a Parts Department newsletter to notify customers and field sales personnel of the availability of new and improved parts, special deals on parts through sales or inventory closeouts, or of matters of policy pertaining to the sales and marketing of parts. The publication was introduced in the middle of 1953 and issues appeared at irregular intervals for several decades. The name “Inside EMD” was subsequently adapted in the 1980s for the internal company newspaper.
The original "Inside EMD," printed in the 1950s and 1960s, was one of the most basic of the EMD internal commercial publications. It was printed on inexpensive paper and seldom included any photographs. However it frequently contained technical information that was not conveyed in any other publications.
One small item that I recall seeing during my visit to the EMD Ogden warehouse was that the first GP9 to receive 48-inch radiator fans was order number 5552-21. I wrote the info on a scrap of paper and headed home. Using my recently acquired copy of the EMD 1972 product reference, I looked up order number 5552. I found that it was for UP 300-349 and 300B-349B, built in July to October 1957, and that 5552-21 was UP 320, being the 21st unit of order 5552. Later research found that EMD regularly made arrangements to retrofit previous units in a particular order to the latest configuration, meaning that UP 300-319 likely received 48-inch fans after leaving the factory. I had not seen a builder's photo of any of these unit to verify that assumption that UP 321 was the first.
Back in May 1999, while doing research among photographs of Union Pacific diesel locomotives, I noticed that the 300s were delivered with 48-inch radiator fans. That break between 321 (built in July 1957) and 322 (built in September 1957) was where I thought the break was between those delivered with 36-inch fans and 48-inch fans, according to that EMD publication I saw 25 years before, making me think that the units delivered with 36-inch fans were retrofitted within six months with 48-inch fans (I mentioned this in a photo caption for the photo of UP 300, taken in April 1958, in my article in Diesel Era about UP's turbocharged GP9s). I discovered that this was wrong after looking again at Harold Ranks' photo of UP 306 being delivered in July 1957. The photo shows the unit with 48-inch radiator fans, and, by the way, also showing that the 300s were the first units with aluminum paint on their trucks.
I've often wished that I had paid better attention to those books in the EMD warehouse. They closed their Ogden warehouse in about 1978, when they opened their warehouse in Commerce, California.
Regardless of which radiator fans were used, a feature also discovered in 1999 was that UP 300-321 had the single large louver set on the battery box doors, and 322-349 (and 300B-349B) have the two small groups of three louvers, like the later GP18s and GP20s.
Installation of the RCS (Radio Control Systems) radio equipment was in the short hood of the SD45s. This necessitated the relocation of the Coded Cab Signal equipment box to the conductor's side walkway, behind the cab.
The initial antenna configuration consisted of a pair of "firecracker" antennas mounted directly on the locomotive cab roof. Difficulties in radio signal reception resulted in a redesign of the antenna system to include an antenna platform, or ground plane, mounted on risers above the cab roof, with a pair of can style antennas mounted on the platform.
The normal voice communication firecracker antenna was relocated to the back of the long hood, between the curved grab iron behind the rear most radiator fan, and the sand filler hatch.
Measurements of the RCS antenna ground plane platform, were furnished by Jim Booth Jr., in an email dated May 24, 2000.
- The RCS antenna platform was 58-1/4 inches by 64 inches.
- The can antennas were 30-5/8 inches from front edge and 9-1/2 inches from side edge.
- From top of table to the cab roof surface was 15-1/4 inches.
- On one version of the RCS platform, the platform itself was about 2 inches thick, using 2 inches by 2 inches angle on front and back edge, with 8 braces made from the same ange iron equally spaced on the underside.
- On the tapered version of the platform, the thickness at the sides was 3-1/2 inches in the center, tapering out to the front and rear edges, which were 2 inches thick.
- The legs were 3 inches by 3 inches angle iron.
At Salt Lake City, the shop crews had to cut the platform off to lift the units with the 250-ton crane, because the platform interfered with the cross piece of the crane. The weld lines for these legs, about 3 or 4 inches above the roof, can be seen upon close examination of photos of the locomotives.
Roof Top Flashers
Dick Harley wrote about UP's use of roof top flashers to the UP Modelers Yahoo group on November 4, 2011:
In searching the 'ST' drawing index between late 1965 and early 1969 (drawings ST-8400 to ST-8800) I found installation drawings for a "Rotating Warning Light" listed for E8/E9 and U50 locos issued in March, 1967 and for SD40 locos issued in April, 1967. A drawing for "Dual Rotating Warning Lights" for SD24 #423 (SD24M #3100) was issued in April 1968. In November, 1968 drawings were issued for wiring diagrams for rotating warning lights on GP35, DD35A, SDP35, SD40 and SD45 locos; and for installation of a warning light to GP30s. And, in January, 1969 a drawing was issued for installation of a rotating warning light on EMD switchers. Whether the "Used On" classes on any of those drawings were expanded to other classes by revision after original issue, I cannot tell. I have none of those drawings, nor any of the correspondence that would have accompanied them to the shops.
The only Operating Rules book that I have found, so far, to mention those warning lights is the May 1, 1972 Operating Rules book for the UP RR Company. It states in Rule 17 (F) the following: "Revolving amber lights on locomotives so equipped must be burning both day and night as follows: On road engines, when engine is moving, except on trailing units in multiple consist. Light must be extinguished when stopped clear of main track to meet a train. On yard engines, when moving in a street and when approaching or passing over public or private crossings."
I have found no correspondence so far that discusses the whys or how effective those lights were. I have not documented what units had warning beacons when new, though I know DDA40Xs did.
Union Pacific issued a separate letter with each locomotive order, listing the specifications showing how each order would be equipped from the factory, and what UP would add at Omaha during the set-up. The letter for GP9s 300-349 in April 1957 does not mention a roof warning light. There is no mention of roof top warning light in the January 1968 letter for the SD45s, 3600-3649. In November 1968 a drawing was created showing the installation of rotating warning lights on GP30s 700-735 and 800 and 801 (ST-8778).
A letter for DDA40X 6900-6924, dated March 19, 1969, shows "Apply one Western D-312 amber warning light on centerline cab roof (Mounting bracket and provision for wiring applied by EMD)."
Drawings were issued in October and November 1971 showing the application of Western D-312 rotating warning lights on the remaining GP9s, SD7s, GP20s, and Alco C-630s (ST-9029, 9038, 9041).
Apparently, the Western Model D-312 revolving light was used until some time in third quarter 1974. It was called out specifically as late as the GE U30Cs 2870-2904, letter dated January 24, 1974, "Warning Light, Western D-312". The letter for the GP38-2s, 2040-2059, dated September 27, 1974, only mentions that EMD was to provide "provisions for application of warning light to cab roof."
The letter for EMD SD40-2s 3288-3304, dated November 4, 1974, showed "Warning lights - Reflectolite Model 6551", by authority of a letter from F.D. Acord dated August 13, 1974 (FDA-296). A similar letter, dated December 3, 1974, for the U30Cs 2905-2919, shows the Reflectolite Model 6551, by authority of the same letter from Acord (FDA-296). The letter for SD40-2s 3335-3409, dated January 11, 1977, continued to show the warning lights as the Reflectolite Model 6551.
(A review of the engineering drawing index shows that no drawings were issued covering these Reflectolite warning lights. Also, no drawings were issued after those in 1971 covering the Western D-312 rotating warning lights.) (The index ends with drawing ST-9797, dated October 1978.)
In response to concerns for increased crew safety, Union Pacific and General Electric worked closely to design a version of what has been called the North American Safety Cab, a concept already well developed in Canada. The result, completed in November 1988, was GE's wide-nose demonstrator B39-8W no. 809 (former B39-8 808, and originally built as B36-8 demonstrator 606). GE no. 809 was sent to each of the railroad's major terminals for crew evaluations. Union Pacific reaction to the design of the cab on the 809 was mixed, with many details in need of improvement. General Electric, with a great deal of input from Union Pacific operating department personnel, went back to the drawing board and came up with the new design that made its first appearance on UP 9356 in December 1989.
In an article in an August 1994 issue of the Journal of Commerce, Robert Grimaila, UP's general director of maintenance planning and technology, remarked that the new safety cab provided "additional functionality. By extending the full-width cab forward over the locomotive nose, the crew has an additional 24 square feet of work space. The extra space is used to relocate what used to clutter the old cabs. Electronic and control equipment that would have to be hidden beneath the floor is more accessible and serviceable. In addition to safety features for the crew, and space to declutter the crew space, we have added insulation that has allowed us to make the first stabs at really reliable heating-ventilating-air conditioning systems." The new safety cabs added about $50,000 to the $1.5 million cost of a new locomotive.
UP's first safety cab units were SD60M 6085-6268, delivered in January 1989. These 184 units, delivered in January 1989 through November 1990, were fitted with the earlier version of the North American Safety Cab with three cab-front windows and straight nose sides.
Union Pacific's first order of wide-nose General Electric Dash 8s (also called "Super Cabs") came in December 1989 through March 1990, with the delivery of 50 units carrying road numbers 9356-9405.
One of the design differences between the GE and EMD versions is GE's use of sloping front cab windows to reduce glare. Another improvement, made on both GE and EMD units, is the nose mounted headlight, moved from above the windshields, to eliminate the glare of the headlights off of the top of the low nose at night.
Later changes to the safety cab design on EMD locomotives included a small change to the slope of the nose at the outside corner, lowering the corner a small amount to allow a person riding on the front steps to be seen from inside the cab. A more major change was from three front windows to two front windows. A summary by Sean Graham-White shows the changes to the EMD cab on UP:
- UP 6085-6215 were the first three window design (delivered January 1989 to June 1989).
- UP 6216-6268 had the same three window design but had the slight dip at the outward edges at the front of the nose (delivered September 1990 to November 1990).
- UP 6269-6365 had the two window cab and angled inward "cheeks" at the front sides of the nose (delivered November 1991 to October 1992).
This later version of the safety cab (two windows and inward-angled nose sides) remained in production by EMD from late 1991 through early 1999 with the delivery of the last SD9043AC units numbered as UP 8179-8308, delivered in January 1998 to January 1999. The first SD70Ms on UP, beginning in April 2000 with UP 4000, and continuing through UP 5126, delivered in November 2002, were also equipped with this later version of the two-window safety cab.
Beginning with UP 5127 in September 2004, UP's SD70 units, nicknamed as "new cab SD70s" came equipped with all-new design cab, greatly changed from the earlier designs. UP's newest SD70ACe units, as of late 2013, continue to use this new design safety cab.
For the GE units equipped with safety cabs, the design has essentially remained unchanged from the earliest C40-8W (C41-8) of 1989, through to the newest ES44AC (C45AH) units delivered in late 2013, numbered up to UP 8111.
Nose Door Windows
Originally, windows were installed on the nose doors of safety cab units to allow a crew member leaving the cab to see a fellow crew member on the walkway prior to opening the door. In practice, however, on-coming crew members nearly always stayed on the ground until the off-going crew climbed off. It is also pretty difficult for two people with their grips and other necessary items, to pass each other on the front platform; there simply isn't room.
The first SD60Ms on UP to receive nose door windows from the factory were UP 6216-6268, delivered from September to November 1990. The earlier units, UP 6085-6215, were delivered in December 1988 to June 1989 without nose door windows, and the feature was added later by UP. UP's first Dash 8-40CWs began arriving in December 1989, numbered as UP 9356-9405, and were equipped with nose door windows right from the initial delivery.
A project to remove the nose door windows began during early 1999 because crews began to be concerned about foreign objects and material from a grade crossing collision entering the cab area through a broken nose door window. The most apparent units to have their nose door windows removed are the SD60Ms that are renumbered into the new 2240-2520 number series.
Snow Shields on E8/9 Units
When and why did UP put snowshields on top of their E8s and E9s?
On Union Pacific, by early 1955, the E8s and E9s were experiencing cooling problems, and operating officials determined that the small opening on the side behind grilles was restricting the amount of intake air available for the engine. The road's solution was to cut an opening in the top of the winterization hatch, immediately above the reverse-mounted intake fan. Tests of this new configuration revealed that the new top opening allowed rain and snow to be pulled into the engine room, causing electrical grounds, since the electrical cabinet was located right below the intake filter box. UP's solution to this rain and snow problem was the road's trademark "snowshields," mounted above the extra opening in the winterization hatch. Snowshields were installed on the road's E8 and E9 fleet, beginning in 1955-1956, and continuing through the late 1950s.
The original question was asked on the Diesel Modeler's email group in July 1999: "UP was well known for applying snow shields over the winterization hatches on its E8s and E9s. However, they were applied over the blank end of said hatch, not over the grill end, and I am wondering just what they were supposed to do. Can anyone explain how they worked? I have applied these shields (Utah Pacific Part #83) to my UP P2K E8A and kitbashed E8B, but am wondering about their intended function."
In July 1999, Steve Orth asked Bill Metzger for his comments about whether or not there is a hole under the snow shield. Bill answered on July 21, 1999:
"Yes, there were holes cut in all winterization hatches. This was to improve airflow in the thin air of Wyoming. Short of turbos, this was the best way to get more air into the carbody (remember the GP9 experiments). E-8/9 drawings show that the "front" fan at both ends point down (into car body). They pulled air into the side ducts in the conventional system but it was not enough in the low Oxygen of high altitudes. So UP cut the extra opening on the top directly above the fan to allow a direct shot into the body and supplemented the standard flow. We had a diagram with colored arrows of this all ready to go for the E unit Streamliner articles but it was eliminated to limited color signatures of that era.
So now there was a 36" hole in the roof of the hatch sucking air directly down into the carbody. Therefore the snow shields. I spent an entire day in mechanical in Omaha looking for a diagram to show the modifications and the design dimensions of the shields. No Luck! I even talked to a retiree who said "the diagrams are in the file cabinet over to your left in either the bottom drawer on the on right above it". But no luck--gone home in someone's personal file I guess. The 951 as the Preamble Express had the shields removed because of restricted clearances on it's trip around the country. Several photos were published looking down from bridges that show the "holes" with the chicken wire grill.
And since their run of about 1989, Overland models has included these openings on their models of UP locos."
The air intake on an E8/9 was through the side grilles, via a duct that funneled the intake air up to the roof hatch, then down through the 36 inch intake fan and into the filter compartment. Maybe the intake fan and air filter compartment was EMD's early attempt at a pressurized carbody. After entering the car body interior, the intake air was used as combustion air for the diesel engine itself, and by the generator and traction motor cooling blowers. The combustion air was further filtered by a bank of oil soaked wire mesh filters mounted atop the Roots blowers.
Quoting Preston Cook (writing as W. A. Cuisinier) in Extra 2200 South, Issue 43, Nov-Dec 1973, p. 21:
"E8 air intake for engine and equipment blowers is rather novel. Located in the winterization hatch of each prime mover is a fourth 36 inch fan, similar to the three [radiator] cooling fans, but which blows air downward, into the carbody, through a filter compartment built into the roof hatch. Oil wetted wire mesh cleans the air just before it enters the engine room. This fan can draw air supply from either of two sources, from a duct thru which air enters the carbody behind stainless steel trim grill during summer, or in winter by switching (using handle in the engine room) the position of flaps inside the winterization hatch, it can draw warm air from the exhaust stream of the #1 radiator cooling fan. Unfortunately, like many other devices, this air system too can be completely circumvented by the same negligence, failure to replace the false ceiling panels."
He continues describing the dangers of having unfiltered air in the carbody, which can cause road failures by flashovers due to dirt-clogged electrical gear. There is also an excellent drawing that shows the flow of air through the roof top hatch, in both summer and winter positions. For those who are close, or who plan to be there on either vacation or business, the E8 model at the Smithsonian in Washington DC also shows this in section and cut-away.
This issue of X2200 also has several detail photos of the carbody interior of E8/9s, along with a cut-away drawing, and a complete E8 and E9 roster, including Amtrak, as of March 1, 1974.
A completely different reason for the snow shields is told among UP's mechanical employees. Warren Johnson, a 30-year veteran mechanic who retired from UP's Salt Lake diesel shop, remembers being told that the snow shields on UP's E units were in fact "put on top of the locomotive to disturb the air at the top of the locomotive. The reason they wanted to do this was to break up the exhaust coming out of the top of the locomotive. Who would want to ride in a domeliner and see nothing but exhaust? It worked the same as the smokelifters on the steam engines used in passenger service. The original pieces had angle iron welded to the top as a stiffener as well as something to disturb the air flow across the top."
Snow Shields on GP9s
On the GP units, the general intake air used by the electrical gear was filtered by the filters on the carbody doors, behind the louvers. As with the E8/9s, the engine combustion air was also further filtered by a bank of four standard 18 inch square filters mounted atop the Roots blower. The winterization hatch on a GP unit was not part of the intake air system as on the E8/9. On the GP units the hatch simply allowed heated radiator exhaust air to enter the carbody space above the front of the diesel engine (which was towards the rear of the locomotive). The lever to change from summer to winter was located on the hatch itself, and was held in either position by a bolt. This required the position to be changed by climbing up on the top of the locomotive.
The structures on top of UP's 12 AiResearch turbocharged GP9s modified in 1955-1956, and in 1959, are definitely snowshields. On the AiResearch units, their sole source of intake air was on top of the units, through a filter box that contained eight of the standard 18 inch square oil soaked wire mesh filters. The drawings reproduced in the July 1988 issue of the UPHS "The Streamliner" magazine show this.
The carbody louvers would not have been blanked off, since open louvers would furnish the needed air for the generator cooling fan, and the two rear traction motor blowers. (This brings to mind that the two louvers on the equipment door under the GP7 cab were there to provide air for the two front traction motor blowers. EMD soon found out that they were not needed, so to save production costs on later GP9s, GP18s and GP20s, a blank door was used instead.)
As part of UP's GP9 turborcharging program in 1955-1956, additional holes in two GP9s were done to address operating temperature problems at the high altitudes of Wyoming and other western locations on UP. During the initial turbocharger tests in southern California, the modified units ran too hot (on Cajon Pass in California, not in Wyoming), and UP cut the holes to improve air flow to help them run cooler. The holes did not help, and the units ran at almost the same temperature. This is from Lloyd Edson, the head of the turbocharging program. Like on the GP9s, the hole in the top of the winterization hatch on E8/9s may have been added to help the units run cooler. The two programs would have been done during the same 1955/1956 time frame.
UP 3794 was fitted with rear toilet enclosure in February 1982 through November 1990. The modification was done as a test to alleviate crew complaints about odor from toilets installed in the low noses of UP's other units, but was unpopular with the crews due to safety concerns while accessing the new lo action during inclement weather at track speed. To save maintenance costs, the enclosure's access door was welded shut by mid 1985, and was removed completely in November 1990 when 3794 was rebuilt as one of the prototype units of the SD40-2 Life Extension Program, which rebuilt as total of 423 units between late 1990 and late 1999.