THE TUNNEL MOTOR - and EMD's Extended Fight with Thermodynamics

 

    Southern Pacific had a unique problem in the late 1960s and early 1970s. While most other roads had success operating EMD's latest line of road locomotives (at the time, the 645 line), SP was running into trouble on its mountain routes. EMD's SD45 (of which SP had over 300) were overheating in tunnels and snow sheds at an alarming rate. Even stranger still, GE's new U33C was not having any issues at all with overheating and was operating perfectly fine through the road's many bores and sheds. What was going on here? SP embarked on a quest to find out, and so began the saga of the "tunnel motor", or EMD's never ending quest to meet SP's demanding operating conditions.

    First, an extremely simplified discussion of thermodynamics and how EMD radiators perform compared to GE radiators. Pictured here will be an SD45 and U33C, however the following description applies to almost all of EMD's GP/SD line and GE's Universal line with a few notable exceptions (covered later on in this post). 

    EMD's SD45 draws cool radiator air from around the shuttered intakes at the rear of the unit. When extra cooling is necessary, the radiator automatically regulates itself using the radiator shutters and fans. In the case of the SD45, two banks of three radiator cores with two separate sets of shutters and three distinct fans are present. One fan activates, then one set of shutters, then one fan, then the other shutters, then the third fan, when necessary. When little cooling is necessary, the two rear fans are used as an air intake while the shutters remain closed. At maximum notch, full load, the engine can sustain itself with all three fans running and both shutters open.

    GE's U33C has one central fan and six radiator cores. Air is drawn from vents on the side of the hood, below the radiator, and drawn upward over the cores. Mounted directly at the top of the cabinet, the radiator fan pushes air over the cores upward. Using a fluid valve to regulate water flow, the radiator allows more or less water through depending on necessary cooling. At maximum notch, full load, the engine sustains itself with the fluid valve completely open and the fan at max speed. This eliminates the need for shutters as well as requires only one engine driven fan (compared to EMD's three electric fans.) 

    One last notable bit of engine tech is the exhaust. On both the SD45 and U33C, a single exhaust stack expels engine exhaust directly upwards. On the SD45's 20-645E3, this stack is positioned on the generator end of the engine (closer to the cab), while on the U33C's 7FDL16E the stack is on the auxiliary end (adjacent to the radiator).

Relative heat diagrams for the U33C and SD45 in an enclosed space.
Notice how the U33C's intake is much lower, and gathers cooler air.

 

    The trouble begins under load in tunnels. Operating in open air, both the U33C and SD45 work as intended; both units expel hot air upward and intake cold air from the engine air intakes without issue. However, when inside a tunnel, it was found that SD45s were inhaling hot air as a combination of engine exhaust and hot radiator air due to the close proximity of the radiator intake and fans to the radiator exhaust. This problem was not extended to the Universals, as their lowered cool air intake was able to intake colder, lower air than the hot rising exhaust from the engine and the already spent hot radiator air.

    SP was a firm believer in improving a product as much as possible, and went to EMD with their findings. In response, EMD developed the "Tunnel Cooling Modification" which reoriented the radiator to be more akin to GE's design, placing the cores at the very top of the hood and placing the fans underneath them blowing upward. Depending on installation, a different configuration was used; on the SD45T-2, three fans are present, on the SD40T-2, only two are necessary. On EMD's tunnel cooling modification, in place of the fluid valve regulator present on the Universal, shutters placed above the cores were used, opening as cooling was necessary. 

Builders photo of SSW 9163

Exposed radiator cores behind
open shutters; a solid photo
of how the Tunnel Motor style
radiator was oriented.

    EMD would eventually make changes to their radiators to make them perform better in enclosed spaces. Effective after 1975, baffles between the separate fans and requiring the shutters to open immediately upon fan activation (rather than drawing air through the other fans, which would be significantly hotter than the air around the shutters) were made standard on the Dash 2 line and all future models with the same radiator design, continuing to the present. Tunnel cooling modifications were offered on the 50 Series of the 1980's, however no purchasers optioned it (SP was not interested in the 50 Series and DRGW's one order of SD50's was a duplicate of a Seaboard order).

    The "tunnel motor" radiator seemed to be more than satisfactory to SP, as they purchased 247 SD45T-2's and 239 SD40T-2's from 1972 to 1980, with DRGW also purchasing a handful of SD40T-2's during the mid-1970's (Rio Grande had similar troubles with their EMD products in similar circumstances). Eventually, after merger with DRGW, SP owned all built tunnel motors, operating them with distinction well into the 2000's on UP. Some were exported to South America, others were sold into secondhand service, and a select few were donated to museums (at least one SD45T-2 and two SD40T-2's).

An MP15DC. The radiator intake is clearly
noted at the front face of the hood.
Craig Garver photo

    While SP was dealing with their troubles in tunnels, EMD encountered a similar issue with their smaller power and overheating; however, unlike the road switchers, the switcher and light road switcher units had a completely different radiator design. Originally introduced in the late 1930's on their S and N line switchers, EMD's switcher radiator draws air from the front and exhausts upwards, across the cores.

 

    This presented a similar issue to the road switchers, but in a different context: when two switchers were run front-to-front (usually SW1500's or MP15DC's), the radiators could not draw enough cold air to keep the engine cooled, leading to overheating. This was most exacerbated on roads using them in road service (namely P&LE, among others) where units would overheat each other facing towards themselves. As it turned out, the tunnel cooling modification was called upon here, where a downsized version of what was present on the SD's was used in place of the standard forward facing radiator fan. Found satisfactory, the tunnel cooling style radiator was made standard on the switcher line. The MP15AC was the first switcher to employ this cooling setup in 1975, and all further locomotives of this caliber made by EMD (such as the GP15 and MP15T) would carry the "tunnel cooling" style radiator standard.

An MP15AC. The radiator has been completely
reoriented, however the rest of the unit is
largely the same on the outside.
Craig Garver photo

    GE never changed their radiator setup in any dramatic way, simply iterating upon it and upgrading it with improved components and customer preferred options. Replacing the shaft driven fan with an electrically blown fan, moving the radiator cores upward in the body, and generally simplifying construction were among the many improvements GE made to their radiator design, up until today's Evolution line locomotives. Ironically, despite the roads findings, the Universals were disliked after the mid 70's recession and were operated minimally by an ailing Southern Pacific. While the U33C's of the 1970's were retired early, EMD's SD45T-2's and SD40T-2's had long and legendary service careers on SP, DRGW, and UP, so it's important to remember the key role GE had in the creation of an icon of western railroading in the late 70's and 80's.