Boeing XP-32 (Model 278A)
The Boeing XP-32 was the designation given to a un-built version of the YP-29 that would have been powered by a 700hp Pratt & Whitney engine.
The YP-29 was an improved version of the Boeing P-26 'Peashooter', developed in 1934. The P-26 was the first monoplane fighter to enter USAAC service, but it was a transitional design, with braced wings, a fixed undercarriage and an open cockpit.
The YP-29 used the fuselage and engine of the P-26 but with cantilevered wings that eliminated the external bracing and a retractable undercarriage. Three prototypes were built, two with cockpit canopies and one with an open canopy. They were powered by the Pratt & Whitney R-1340 Wasp engine. The YP-29 wasn't much of an improvement over the P-26 and didn't enter production.
The designation XP-32 was allocated to a 1934 version of the YP-29 that would have been powered by a 700hp Pratt & Whitney R-1525-1 twin-row radial engine. It would have been a low winged monoplane with a enclosed cockpit and tapered wings with a straight leading edge and curved trailing edge. The aircraft was never built but the retractable landing-gear designed for it was later used in the Brewster F2A-1 fighter.
Engine: Pratt & Whitney R-1535-1 Twin Wasp Jr
Length: 27ft 5in
Gross Weight (as designed): 3,895lb
Estimated maximum speed: 250mph at 7,500ft
Progress Rail Locomotives, doing business as Electro-Motive Diesel (EMD) is an American manufacturer of diesel-electric locomotives, locomotive products and diesel engines for the rail industry. The company is owned by Caterpillar through its subsidiary Progress Rail.  
Electro-Motive Diesel traces its roots to the Electro-Motive Engineering Corporation, a designer and marketer of gasoline-electric self-propelled rail cars founded in 1922 and later renamed Electro-Motive Company (EMC). In 1930, General Motors purchased Electro-Motive Company and the Winton Engine Co. and in 1941 expanded EMC's realm to locomotive engine manufacturing as Electro-Motive Division (EMD).
In 2005, GM sold EMD to Greenbriar Equity Group and Berkshire Partners, which formed Electro-Motive Diesel to facilitate the purchase. In 2010, Progress Rail completed the purchase of Electro-Motive Diesel from Greenbriar, Berkshire, and others.
EMD's headquarters, engineering facilities and parts manufacturing operations are based in McCook, Illinois, [note 1] while its final locomotive assembly line is located in Muncie, Indiana. EMD also operates a traction motor maintenance, rebuild and overhaul facility in San Luis Potosí, Mexico.
As of 2008, EMD employed approximately 3,260 people,  and in 2010 it held approximately 30 percent of the market for diesel-electric locomotives in North America. 
XP-32 was the USAAC designation given to the Boeing Model 278A, a company-financed design project of 1934. The XP-32 was basically a developed version of the earlier P-29 with a 750 hp P & W R-1535 Twin Wasp radial engine. The project drawings show a low-wing, cantilever monoplane design with a fully-retractable undercarriage and a fully-enclosed cockpit with a rearward-sliding canopy. The XP-32 design looked a lot like the Model 264 (YP-29A), but the XP-32 differed in the means by which the undercarriage was retracted. Whereas the P-29's main wheels retracted rearwards to lie partially-exposed underneath the wing, the main wheels of the XP-32 retracted inward to be stowed flush with the sides of the fuselage, a pattern that would be followed by the Brewster F2A-1 Buffalo of 1938. The gross weight was 3895 pounds.
The USAAC did not encourage the development of the project, and the XP-32 never got past the design stage. Boeing got out of the fighter business altogether shortly thereafter. Boeing was not to submit another fighter design to the military until the XF8B-1 long-range carrier-based fighter-bomber of late 1944.
- The American Fighter, Enzo Angellucci and Peter Bowers, Orion Books, 1987.
Boeing XP-32 (Model 278A) - History
As with the competing Lockheed Martin X-35, Boeing designed three variants of the X-32 for evaluation. The conventional takeoff and landing (CTOL) X-32A was developed for the US Air Force, the short takeoff and vertical landing (STOVL) X-32B for the US Marine Corps and UK Royal Navy, and the carrier-based (CV) X-32C for the US Navy. However, only two flying examples were actually built.
The first to fly, the X-32A, was used to demonstrate overall flight characteristics, systems, and control software. This model was also used to evaluate the low-speed handling and carrier-approach qualities of the X-32C naval variant. The second example, the X-32B, was equipped with a direct lift system for STOVL operations and was used primarily to evaluate vertical flight and hover characteristics.
Boeing's strategy for STOVL flight was based on that used in the British Harrier. The single engine was mounted at the center of the vehicle and its thrust directed through three movable nozzles permitting vertical flight. Boeing preferred this approach to Lockheed's lift fan design citing it as less risky. Nevertheless, this hover method was ultimately viewed as a limitation of the X-32 design.
Boeing was also penalized for proposing several changes between the X-32 demonstrator and the final production model. Among these changes was abandonment of the variable intake cowl designed for the X-32C, replacement of the twin tails on the X-32 with more conventional vertical and horizontal tails, and a redesigned wing (see the 3-view below). These factors caused Boeing to lose the JSF contract during the October 2001 downselect, and Lockheed Martin was instead chosen to build a production F-35.
Data below for X-32A
Last modified 26 September 2009
United States (US Air Force)
United States (US Marine Corps)
United States (US Navy)
18 November 1930
18 November 1930: The prototype Boeing XP-9, Air Corps serial number A.C. 28-346, a single-seat, single-engine monoplane pursuit, made its first flight at Wright Field, Ohio.
This was Boeing’s first semi-monocoque aircraft, built of a sheet dural skin over metal formers. The Army Air Corps issued the contract 29 April 1928 and the aircraft was completed in September 1930, then shipped by railroad to the Army test base.
The XP-9 (Boeing Model 96) was a single-place, single-engine high-wing monoplane with fixed landing gear. It was 25 feet, 1.75 inches (7.665 meters) long. with a wingspan of 36 feet, 6 inches (11.125 meters) and height of 7 feet, 10.25 inches (2.394 meters). The prototype’s empty weight was 2,669 pounds (1,211 kilograms) and its maximum takeoff weight was 3,623 pounds (1,643 kilograms).
The pursuit prototype was powered by a pressurized-liquid-cooled, supercharged, 1,570.381-cubic-inch-displacement (25.734 liter) Curtiss Super Conqueror SV-1570-C dual-overhead camshaft (DOHC) 60° V-12 engine with 4 valves per cylinder. This engine was rated at 600 horsepower at 2,400 r.p.m. It weighed 920 pounds (417 kilograms).
The airplane had a maximum speed of 213 miles per hour (343 kilometers per hour). The service ceiling was 26,800 feet (8,169 meters). Armament was a combination of two machine guns, either one .30-caliber and one .50-caliber, or two .50 caliber, mounted one each side of the fuselage, firing forward.
The placement of the single high wing seriously restricted the pilot’s vision, making landings very dangerous. The airplane was highly unstable in flight. Increasing the size of the tail surfaces did little to improve this. After just 15 flight hours, the XP-9 was permanently grounded and was used as an instructional airframe.
The performance and handling of the XP-9 was considered to be so poor that an option to buy five pre-production models was canceled.
The XP-9’s sole redeeming quality was its method of construction, which has been almost universal since that time.
Boeing’s little known 307 Stratoliner, affectionately dubbed "the flying whale" for its portly lines, ushered in a new aviation era when it entered into airline service in mid-1940. It was the first in-service pressurized airplane and airliner. It is cabin pressurization (termed cabin supercharging at the time), along with air conditioning and heating that enables today’s high altitude passenger jet airliner flights above the weather and turbulence, where the thin air and sub-zero cold could kill passengers within minutes were they unprotected. The Seattle-built, propeller driven Stratoliner took the first practical step on the journey to safe high altitude passenger flight. Although only 10 aircraft were built, it was very successful in airline service one was reported still carrying passengers in 1986. Remarkably, at least two airframes survive today, the restored Pan American Airways NC19903 Clipper Flying Cloud, which began flying again on July 11, 2001, and the fuselage of the Howard Hughes special model, which is now a yacht. As luck would have it, the Flying Cloud was the first in-service pressurized airplane and airliner.
The Stratoliner was arguably the most advanced operational aircraft in the early 1940s, for it also utilized power boosted control surfaces and geared two-speed engine superchargers, lacking only the tricycle landing gear employed by the Douglas DC-5 airliner that entered service with Dutch airline KLM a month earlier. Building upon that lead, during World War II the U.S. was the only country to field:
- a pressurized strategic bomber
- pressurized turbojet powered fighters
- a pressurized transport/airliner
- aircraft employing power boosted control surfaces.
Germany and the United Kingdom operationally deployed pressurized propeller-driven bombers and fighters, which were modifications of earlier aircraft. Today’s high performance turbine-engined civil and military aircraft are pressurized and utilize powered flight controls.
1930s High Altitude Flight
In the 1930’s aviation researchers realized that flying at high altitude above the weather would pay dividends in passenger comfort, higher speed, and longer range. Progress had been made to fly safely at high altitude: reliable oxygen masks, electrically heated flying suits, a practical pressure suit, and a successful experimental pressurized airplane that flew in May 1937. During this period, airlines, the military, and individuals conducted high altitude flight tests, which resulted in several U.S. and a British airline requesting proposals for pressurized airliners.
Boeing, Curtiss, and Douglas responded with designs, all of which reached the flying hardware stage by 1940. Britain’s Fairey built a mockup before the project was cancelled in 1939 due to World War II. First in the air, however, was the U.S. private venture Abrams Explorer two-crew photomapping airplane flown during November 1937. The Explorer, the only one built, flew successfully for many years and now resides in the National Air & Space Museum collection in Washington D.C.
Stratoliner Derived from B-17 Bomber
Wellwood Beall, famed 314 Clipper flying boat designer, led a talented team that in December 1935 began development of the 307 as an airliner derivative of the model 299/XB-17 Flying Fortress. Douglas, by 1936, had five airlines sponsoring development of its pressurized DC-4E four-engine long-range landplane airliner. Pan American Airways (PAA) and Trans Continental and Western Airlines (TWA) decided before it flew that they wanted out, due to high costs and projected performance shortfalls. In 1937 they ordered instead the 307, four for PAA, five for TWA. Millionaire Howard Hughes later ordered another. These were to be the total orders for the aircraft, which cost approximately $250,000 delivered. Breda of Italy had sought a production license in 1939, wanting the 307 for transatlantic service and for its technology. Political and Not Invented Here considerations evidently killed the project.
First Flight in 1938
Before the Stratoliner took wing, a confident Boeing designed a huge pressurized two-deck flying boat in response to a 1937 PAA requirement for a flying ocean liner capable of crossing the Atlantic non-stop. Boeing’s model 326 was headline news on June 22, 1938, its announcement coming just 15 days after the 314 Clipper flew. However, Boeing built none of the model 326. Nor were any of four competing designs ever built.
The S-307 NX19901 prototype (for PAA) flew for the first time on December 31, 1938, piloted by Eddie Allen, from Boeing Field, Seattle, for a total of 42 minutes. The first pressurized flight, successfully accomplished by PAA NC19902 Clipper Rainbow, occurred on June 20, 1939.
Design Relationships and Large Vertical Tails
Initially the 307 design was based on the model 299/XB-17 -- wings (over three feet wider), tail and landing gear joined to a new circular section pressurized fuselage with newer B-17B type engines without turbo-superchargers. Production aircraft had wing slots, a dorsal fin, and a large vertical fin -- these last two items were then developed for the B-17E through G models. Large vertical tails have since characterized Boeing airplanes up to the present 777 jet airliner. Development of the B-29 Superfortress pressurization system stemmed from the 307, and this successful bomber was the first mass production pressurized airplane.
The Stratoliner was the first of several Boeing aircraft to use the strato prefix in its name. Strato is derived from the second-from-the-surface of the earth’s atmospheric layers, the stratosphere, which begins at around 30,000 feet of altitude. TWA highflying model SA-307B’s were shy by about 4,000 feet of being able to reach that height. Some early 707 turbojet airliner models were for a time also named Stratoliner -- they cruised comfortably in the stratosphere.
Stratoliner Entered Service in 1940
PAA’s Flying Cloud flew the first operational pressurized service from Miami, Florida, to Latin America beginning on July 4, 1940. The Stratoliner offered unmatched comfort, speed and range advantages over its Douglas DC-3 and Lockheed Electra twin-engine competitors. A wide body airliner, its fuselage was more than three feet wider than the DC-3, and featured a luxurious 33-passenger cabin -- pressurized, air conditioned and heated passenger compartments sleeping berths with windows ample-size individual reclining sleeper seats large seat windows (approx. 12 x 16 inches) men’s and women’s lavatories with skylights and a galley with a skylight.
A month later, a German Junkers Ju 86P two-crew high altitude photo aircraft flew at 41,000 feet over the United Kingdom, becoming the first operational pressurized military airplane.
Boeing Was the 1940s Technology Leader
In 1940 Boeing was king of the hill in advanced technology with its stable of operational airplanes, while Douglas and Lockheed led in sales. With a six-month lead on the 307, the 314 Clipper flew on June 7, 1938. In service with PAA, it was at once the largest, heaviest, longest-range, highest capacity airliner, using the most powerful engines. In the 307 Boeing had the only operational pressurized airplane and the longest-range landplane airliner. Despite having the longer range 314 in service, PAA briefly considered flying the 307 across the north Atlantic, but never did.
During World War II, the Army Air Force (AAF) flew the route with its impressed TWA 307’s beginning in 1942. Boeing’s B-17C was the fastest, highest flying, and longest-range heavy bomber in the air. Adding to this bounty was the XB-29 then being developed, among the most advanced operational airplanes of World War II, with reversible pitch propellers, tricycle landing gear, electronic computer controlled powered gun turrets, and navigation/bombing/tail warning radar systems.
A Special Model for Howard Hughes
A special model SB-307B for Howard Hughes was built with more powerful engines and extra fuel tanks for an around-the-world flight that was cancelled due to the start of World War II. The flight was never made. It was the first Stratoliner delivered to a customer its initial flight (with experimental license NX19904) occurred on July 13, 1939. Postwar it was fitted with a luxury interior, including a bedroom, and named The Flying Penthouse.
A 1964 hurricane severely damaged it and rendered it unflyable. In 1969 it was purchased as scrap for $61.99 -- the fuselage was salvaged (the aft rounded pressure bulkhead formed the cabin after end), then mounted on a boat hull and converted into a luxury yacht named The Londonaire. It was rebuilt beginning in 1994, and is a Florida based, operating yacht named Cosmic Muffin, with N19904 painted on its sides.
World War II Service and Afterwards
During World War II, TWA’s 307’s were taken into AAF service as the camouflaged C-75. PAA aircraft were retained by the airline, with their crews and colors flying under charter for the AAF Air Transport Command. All eight aircraft survived their wartime service.
Postwar Stratoliner airline service began in late 1945 when TWA resumed coast-to-coast flights with its upgraded SA-307B-1 aircraft, and PAA flew the New York City to Bermuda route. PAA Stratoliner service ended in 1948, when its three aircraft were sold. TWA used its five SA-307B-1’s until 1951, after which they were sold off. Three aircraft were in service in Indochina during the 1970s, one aircraft was reported still flying in Laos as late as 1986.
Restored Flying Cloud Airborne
Resplendent in its highly polished aluminum Art Deco finish, the Flying Cloud flew from Boeing Field exactly 61 years and one week after it flew into the history books. The July 11, 2001, flight was actually its third first flight. It flew for the first time in 1939 from Boeing Field. During its active years, the Flying Cloud served not only PAA, but also the President of Haiti and several other owners in its productive lifetime. It made a second first flight on June 4, 1994, to Boeing Field after languishing in the open in the Arizona desert for more than 22 years. Over a seven-year period, a team of Boeing employees and volunteers refurbished it in its Plant II birthplace. Owned by the National Air & Space Museum, the Flying Cloud will be permanently displayed in the museum’s Washington D. C. facility.
[Note: On March 28, 2002, the restored Stratoliner developed engine trouble while on a test flight and ditched into Elliott Bay. No one was injured, and the damaged aircraft was retrieved. The Flying Cloud evidently ran out of fuel, causing it to descend into the water. As of June 2002, Boeing has determined that the necessary repairs to restore the Flying Cloud to flyable condition are cost effective. A team of Boeing and volunteer workers will rebuild it, with the intention of flying east in the summer of 2003 to the Smithsonian Air and Space Museum.]
- First operational airplane with hydraulically boosted control surfaces -- elevators and rudder
- Fastest scheduled long range airline service -- up to 220 mph cruise, flown by TWA model SA-307Bs beginning in 1940
- First airliner (SA-307B) with geared two speed engine superchargers able to cruise at high altitude with passengers in complete comfort beginning in 1940
- First four-engine landplane airliner in U.S. scheduled long-range service
- Wide body fuselage -- wider at 138 inches/11.5 feet overall than its turbojet powered namesake, the 367-80 Dash Eighty 707 prototype tanker/airliner at 132 inches/11.0 feet.
Restored Boeing 307 Stratoliner Clipper Flying Cloud landing at Boeing Field after its third first flight, July 11, 2001
The prototype Boeing 307 Stratoliner first flew on December 31, 1938
On this Day in Aviation History – Boeing XP-9
1930 The Boeing XP-9 monoplane fighter makes its first flight in Dayton, Ohio.
The Boeing XP-9 (company Model 96) was the first monoplane fighter aircraft produced by the United States aircraft manufacturing company Boeing. It incorporated sophisticated structural refinements that were influential in later Boeing designs.
Designed in 1928 to meet the requirements of a US Army request for a monoplane fighter. Its primary contribution to aircraft design was its semi-monocoque construction, which would become a standard for future aircraft. Boeing employed the structural features of the XP-9 into their contemporary P-12 biplane fighter when the P-12E variant incorporated a semi-monocoque metal fuselage structure similar to that of the XP-9. The undercarriage arrangement of the P-12C had also been first tried out on the XP-9 and then transferred into the production model. Only one prototype was ever produced, with the program being canceled because of poor pilot visibility.
Vintage Diesel Design
This week’s column is by Jay Boggess. Next week we will return to the Delta Municipal Power Plant for Part II.
Pretty quickly, early on – when it comes to diesel engines, you hear the word “Roots Blower”. But who IS Roots? Today in the era of Wikipedia, this is an easy question to answer, but not when I was a kid.
I’d first heard of the “GMC Roots Blower” associated with supercharged dragsters & hot rods. Later, while reading my father’s 1944 textbook “Internal Combustion Engines – Analysis & Practice”, I discovered a cutaway section of the General Motors 2-stoke CI (compression ignition or diesel) engine, below:
Later, I learned that Cleveland Diesel, Fairbanks-Morse and Electro Motive Division diesel engines all had Roots Blowers, but no one ever explained why it was called the Roots Blower.
In 2003, a random visit to the History Colorado Museum in Denver came across this artifact:
A mine ventilation blower for ventilating underground hard-rock mines, built by the P.H. & F.M. Roots Company, Connersville, Indiana. The placard listed a date, but the low-res digital pics of the era do not allow me to zoom in – other sources point to the mid 1880’s or so.
Another datapoint came from another random visit, this time to the nearly preserved Bethlehem Steel blast furnaces in Bethlehem, PA (thanks to my former EMD colleague Mark Duve, who insisted we stop).
The building in the foreground of the photo was unlocked, we ventured inside and discovered these:
Very distinctive, two-lobed Roots Blower rotors – look carefully and you will see counter-weighted steam engine eccentrics on the end of the rotors. Inside the same building were the matching horizontal steam engine cylinders for driving these rotors (I took photos but the passage of 16 years has lost those). I later learned that blast furnace blast supply was one of the first uses of Roots Blowers.
So who were P.H. & F.M. Roots? Wikipedia points to a 1931 book, “Indiana One Hundred And Fifty Years of American Development” which provides most of the answers. Philander Higley and Francis Marion Roots were brothers. Francis was the youngest brother, born in 1824, went searching for gold in California in 1849, came home in 1850 and started working with his brother Philander in manufacturing. They patented the “Roots Positive Blast Blower” in 1866. Francis passed away in 1889, Philander passed in 1879. Their company was purchased by Dresser Industries in 1931, and renamed the Roots-Connersville Blower Company. In WWII, they produced low-pressure blowers for blowing ballast tanks in U.S. Submarines, as well as centrifugal blowers for various low-pressure/ high-volume uses, eventually submerged in the vast Dresser product line.
Roots Blower Applications:
Submarine Ballast Tank Blower:
This is listed on the drawing as a 1600 CFM blower, designed and built by the Roots-Connersville Blower Corporation, Connersville, Indiana. The driving motor is a 1750 RPM, 90 horsepower, intermittent-duty DC motor.
To digress extensively – WWII submarines had two systems to blow their ballast tanks – 3000-PSI stored compressed air reduced down to 600 PSI to start the surfacing process and 10-PSI low pressure air supplied by blowers to finish the job once a submarine surfaced. It was this low-pressure job that either Roots Blowers or centrifugal blowers were utilized. Another interesting use was that when a sub is submerged, various tanks are vented inboard the sub, raising the internal pressure of the boat several PSI above atmospheric pressure. If the hatch were immediately opened, the rush of air was known to launch sailors overboard. Instead, the hatch between the conning tower and control room would be shut, the boat surfaced and the bridge hatch opened. While the captain checked to see if the coast was clear, the low-pressure blower is started finishing the blow of the ballast tanks and reducing the excess air pressure inside the rest of the boat.
Fairbanks-Morse Opposed Piston 38D Engine:
The WWII era FM 38D manual does not use the word “Roots Blower” but instead refers to it as a “Scavenging Air Blower”. The FM 38D blower spins at 1450 rpm and provides 6000 CFM at about 2 to 4 PSI. The Direct Reversing version of this engine used a set of linkage and air valves on the blower in order to direct the air in the proper direction when the engine is running astern, thus the blower is running backwards.
General Motors Cleveland Diesel Engine Division (CDED) 278A Marine Diesel:
Cleveland Diesel Engine Division Diagrams – Click for larger
Cleveland Diesel mounted their single Roots Blower on the front of their engine, essentially shortening or lengthening the blower to fit the air flow of the 6, 8, 12- or 16-cylinder models of the 278A, as the photos and following table illustrates.
16-278A – 1700 HP Destroyer Escort Engine: 1650 RPM, 6.5” Hg, 5630 CFM
12-278A – 875 BHP Army Tug Engine: 1650 RPM, 5.5” Hg, 4380 CFM
8-278A(NM) – 800 HP Non-Magnetic Minesweeper Engine: 1833 RPM, 6.5” Hg, 2950 CFM
6-278A – 480 HP 720 RPM Tug Engine: 1358 RPM, 4.5” Hg, 2180 CFM
Thanks to Scott Zelinka for the above Cleveland photos showing a pair of the Spiral rotors used by CDED. The clearances between the rotors is set at .024″ (on the 12 and 16 Cyl) and .018″ on the smaller engines. I find it downright amazing that something with this complex of a shape – and interlocking none the less, could be machined so exacting by hand, and mass produced at that, long before computers and CNC.
With the new Cleveland Diesel 498 engine, a small Roots blower was used in conjunction with the exhaust driven turbocharger to provide for lower RPM scavenging. EMD would solve this issue with their own turbocharger on the 567. A centrifugal clutch drives the blower off of the timing gears that would disengage at a certain RPM and allow the turbocharger to freewheel.
Cleveland 498 diagram
EMD 567/645 Roots Blown Engines
Electro-Motive answered the Roots Blower question in a totally different way than its GM sister division CDED. EMD also had four different engines to support: 6 – 8 – 12 – 16 cylinders. EMD picked one design of blower, then used that one blower for the 6 and 8 cylinders model and a pair of blowers for the 12 and 16 cylinders, changing the blower gear ratio (and blower RPM) between 6 and 8, and 12 and 16 engines, gaining economics of scale and fewer replacement parts to support.
Below is the 8-cylinder 567 model:
Click for larger – Cleveland Diesel engine manual photo – WWII Army ST tug – collection of Jay Boggess
And here is the mid-1950’s 16-567C model. Note the directional air intake, a sign that this engine was likely built for stationary power generation.
Click for larger – Cleveland Diesel Engine Division Photo – Jay Boggess Collection
The 16-567C pic illustrates another clever design feature that EMD incorporated. By placing the Roots Blowers high above the crankshaft (driven by the engine’s camshaft drives), EMD designers provided a niche for a generator underneath the blowers, saving overall length of the engine/generator and thus overall length of the locomotive.
These are just a few short uses of the Roots Blower – several other manufacturers have used them, and coming in one of the next parts on the Delta Municipal Power Plant, we will see a giant Roots-Connersville centrifugal blower used to feed the big 31A18 engine. Roots Blowers are common on many different industrial uses outside of engines.
While many thousands of Roots Blowers have been built, I believe their day in the sun has passed. From my days at the Alaska Railroad, EPA emissions regulations were starting to close in on the Roots Blown engine. I do not know the specifics, but the GP38-2s AkRR owned had to be de-tuned for better emissions, which gave lower fuel economy. And even then, the EPA wasn’t very happy about it (that is, the EPA Tier 0/1/2/3 regulations only allowed de-tuning for existing engines and would not be applicable to a new Roots-blown EMD engine).
So, when you hear an older EMD go by, be it a GP7 or GP9 or 38, think of Philander Higley and Francis Marion Roots and what they invented 150 years ago.
Sidebar – Roots Blower Or Roots Supercharger?
Blogmaster Paul Strubeck has uncovered somewhat heated discussions between the terms “Roots Blower” and “Roots Supercharger”. Both terms can be correct – I will attempt to clarify, but I will preface my comments that I am an electrical engineer by training / experience and only an “armchair” engine guy (from hanging around my father and the many, many gear-heads at Electro-Motive over 22 years).
Supercharging is defined as jamming more air than atmospheric pressure into each cylinder before compression by the piston begins. My 1944 internal combustion textbook notes by providing some form of air pump, you can get more power for the same engine weight or thin-air compensation for an aircraft engine at high altitude.
In the two-cycle diesel engines (FM, Detroit Diesel, CDED, EMD), the Roots Blower acts primarily to scavenge exhaust gases from the cylinder after each power stroke. If the exhaust valves close before intake ports (in the case of a GM 2-cycle diesel), then some supercharging will take place. But the primary purpose is to get exhaust gases out.
If the air pump is driven by a turbine attached to the exhaust manifold, then the arrangement is termed a turbocharger. The turbocharged EMD 645E3 engine provides 3000 THP in the GP40/SD40, while the Roots-blown 645E engine of the GP38 provides only 2000 THP. The Wright radial engine of the Boeing B-17 of WWII used a turbo-supercharger so that it could fly at 25,000 feet over Germany, with each engine producing 750 HP at altitude.
Barney Navarro was the first hot rodder to put a Roots Blower with Detroit Diesel history on a car engine in the 1950’s. The blower, from a Detroit Diesel 3-71 was belt driven off of the crankshaft and made 16PSI of boost in the engine. After that the doors opened and the Roots style blower became a choice power added for race cars (typically drag cars). Today, they are still referred to an x-71 style (in different sizes, including a 14-71, an engine never made), however they are specific made for the application, and not WWII surplus! Supercharging on gasoline/car engines is a much larger topic that literally has had books written on it.
A 14-71 Roots blower on a Pro-Mod car. These blowers are overdriven (the blower turns faster then the crankshaft) to force as much air in as possible.
A little more on a Top Fuel engine – 11,000HP for 3.7 seconds at a time. https://www.hotrod.com/articles/ccrp-1009-8000hp-top-fuel-engine/
Thanks to Jay for writing this weeks post (with some added commentary from me, namely on the Roots Blowers on race cars).
DO-178C and DO-178B Software Certification
The level of effort to comply with the objectives of DO-178 will vary based on software criticality (depending on how software can contribute to a failure condition). The level of effort is also proportional to the size of the software under consideration. DO-178 defines five software levels, each related directly to the failure condition that can result from anomalous behavior of the software. The software level definitions given in DO-178 and the number of objectives required to satisfy the requirements of each level are shown below.
DO-178C software levels
|Failure Condition||Software Level||Number of Objectives|
|Hazardous / Severe – Major||Level B||69|
|No Effect||Level E||0|
The software life cycle data required by DO-178 includes the following:
Plan for Software Aspects of Certification
Provides the Certification Authorities an overview of the means of compliance and insight into the planning aspects for delivery of the product.
Software Quality Assurance Plan
Defines the SQA process and activities.
Software Configuration Management Plan
Defines the CM system and change control process.
Software Development Plan, Software Requirements Standard, Software Design Standard, Software Coding Standard
Defines the processes used for requirements analysis, development, and test for the software product. Includes the standards for requirements, design, and code.
Software Verification Plan
Defines the test philosophy, test methods and approach to be used to verify the software product.
Software Test Plan
Documents the project-specific approach to verifying the software product.
Software Requirements Specification
Defines the high-level requirements applicable to the certifiable software, including the derived requirements.
Tool Requirements Document
Defines the required functional behavior of a verification tool under normal operating conditions.
Software Design Document
Describes the design of the certifiable software.
Software Configuration Index
Identifies the components of the certifiable software with version information necessary to support regeneration of the product.
Software Life Cycle Environment Configuration Index
Identifies the tools end environment used to build and test certifiable software.
Tool Qualification Document
Documents the qualification evidence for any DO-178 verification tools against the requirements established in the PSAC and Tool Requirements Document.
Software Development Folder
Note that this is provided as a set of files on electronic media image using Verocel’s VeroTrace tool. They may not necessarily be maintained as a hard-copy folder. However, traceability between all artifacts still needs to exist and be proven.
Software Development Folder includes as a minimum:
(a) Reference to the applicable requirements
(b) Reference to the implementation (Design & Code)
(c) Evidence of reviews for the Requirements, Design, Code, and Test procedures and test results
(d) Software Test Procedures
(f) Analysis documents for verification, coverage analysis, and any special case analysis.
(g) Change History (CM System)
(h) Applicable Problem Reports
Provides traceability from the requirements to the built software to tests for the delivered software product.
Software Accomplishment Summary
Documents the actual versus planned (per PSAC) activities and results for the project. Provides a summary of the means of compliance used for the software. Justifies any deviations from the plans.
Provides the Source files for:
1. Certifiable software
2. Test Procedures
3. Build and Test Scripts
Documents the results of the functional and structural coverage testing. This includes the actual results and any applicable analyses performed including coverage analysis.
Linkable versions of the “as tested” software.
Addressing the planning, requirements and verification processes
The planning process begins with the Plan for Aspects of Software Certification (PSAC). The PSAC describes the scope system and software that will be considered for certification. The PSAC also describes the overall software life cycle, the software development plan, the software verification plan, the standards that will be used along with the Software Configuration Management Plan and Software Quality Assurance Plan. These core plans and standards define the framework of how the software will be developed and verified along with the transition criteria for each life cycle phase. The PSAC should also call out any tools that will be used to support development and verification processes and identify whether those tools need to be qualified because they automate a process of the software life cycle.
For Level A, B and C software both high-level and low-level requirements need to be developed and verified for the software. For Level D software, only high-level requirements are developed and verified. The verification activities (for levels A-C) should include reviews of requirements, design, code, test cases, test results and coverage analysis. Verocel’s tools such as VeroTrace can manage and control all life cycle data (including reviews and test procedures and results). Verocel’s VerOCode and VeroSource tools are qualified to DO-330, TQL-5 and help with object code and source code coverage respectively.
Previously developed software and DO-178
It is common for applicants to take an existing set of functional software through the DO-178 certification process rather than develop software in a waterfall model while producing the certification artifacts. DO-178 is written as if a waterfall model of development is used, but any software development model could apply. The same objectives apply to software that is engineered to meet DO-178 as software that is developed to meet DO-178 from the start. As shown above, there are many planning documents produced under DO-178. Planning involves not only a strategy for certification, but also a detailed implementation of how the certification will take place including software quality, configuration management, requirements, design and coding standards, and a detailed plan of how the software will be verified. This planning activity involves the entire engineering and quality team and may likely take a number of months to complete.
One way to carry out a process is to use a software prototype (assuming the prototype under consideration is functional and operational) and to capture a set of complete requirements that can be used for test and verification purposes. Normally, these requirements would include both high-level and low-level requirements that map to specific functions in the source code. These requirements would need to be reviewed (independently for Level A) and then used as a basis to construct the test cases for the software. However, before the testing process can begin, the detailed design should be either extracted from the existing source code or developed after the high-level requirements are approved.
The design can be extracted from comments in the source code and put into a descriptive textual document (that shows compatibility with high-level requirements). However, this low-level design information is usually not sufficient to reflect the entire software design. There also needs to be a high-level design document that describes how all the software components will work together, their interdependencies and timing relationships, etc. The low-level and high-level designs also need to be reviewed for accuracy and consistency among other objectives. Once the requirements and design are complete and approved, efforts to review the source implementation and production of test cases can begin. It should be noted that these efforts (requirements, design, code and test) can all happen in parallel provided sufficient configuration management of any approved artifact is in place.
The cost of DO-178 certification varies greatly depending on engineering expertise and code size. One has often heard of the “multi-million dollar” answer when asking the cost of certification of any software to DO-178, regardless of size.
So how to more accurately predict the effort required?
There are a number of ways to scope the level of effort required for DO-178 certification. One useful way to scope the effort is to examine the size of artifacts from previous certification efforts. Consider the example of an operating system certified to Level A: the source code consists of about 12,000 lines of code which resulted in the generation of 1,300 requirements for approximately 700 functions. When determining scope of effort, one useful metric is to plan on 2-4 requirements per function – or 1 requirement for every 5-10 lines of code. Then each requirement will need to be tested, resulting in 1 test procedure for every 2-4 requirements. All requirements, design, code and test artifacts need to be reviewed via checklists. Additionally, the artifacts need to be linked to show traceability between requirements, design, code and tests/results.
What does it cost to certify a line of software to DO-178?
This is a common question asked by software managers. Intuitively, it would seem that it is much costlier to certify software to Level A rather than Level C, given that Level C has only 62 objectives and Level A has 71 objectives to be met. But experience has shown that the difference in cost between Level A and Level C is not that great. This is evidenced by the fact that all the deliverables defined above are required both for Level A and Level C software. In fact, the software planning and software development objectives under DO-178 for Level A and C are identical! The differences between Level A and Level C from a verification perspective impact the requirements, design, testing and analysis processes. Level A verification has the added requirement of independence for some objectives (where the development and verification efforts must be accomplished by different persons). This additional requirement can indeed add engineering labor to the certification effort however, the added cost is likely to be a small percentage of the total certification effort.
Where Level A certification can get expensive is usually through the certification scrutiny an applicant will face when all verification efforts are complete. Convincing an auditor or host of auditors that have the responsibility to “sign off” on software where loss of life could result from failures can be challenging if the applicant has not adequately addressed the planning, requirements and verification processes. Audits can result in added activities that extend project schedules and increase costs. All Level A certification should be planned with a schedule buffer for planned audits and potential rework.
Assessing a cost for all this effort requires an estimation of the engineering labor required for each development and review activity. How long does it take an engineer to produce a requirement from an existing implementation? One way to find the answer is to ask the engineers to estimate the effort for each activity (it is presumed that the engineer is very familiar with the software and design). Thus, a combined estimation can be built based on the engineering labor assessments for requirements, design, tests, reviews, etc.
Experience has shown that the cost of certification per line of code can range from $25 to $100 depending on the cost of labor in a given organization. That would mean that certification of 100,000 lines of code could range from $2.5 million to $10 million. This range should not be applied to any software project however it is only a guideline that must be substantiated and corroborated by the assessment of metrics, a detailed analysis of the source code, engineering judgment, and compensation for risk factors such as certification of object oriented design.
Get in touch!
- A Turbocharged Failure – The Story of the Cleveland 498, Part II June 6, 2021
- A Turbocharged Failure – The Story of the Cleveland 498, Part I May 29, 2021
- When Bad Things Happen I April 28, 2021
- A Roots Blown Experimental Alco? April 8, 2021
- Loss of a Museum Tug – Pegasus March 28, 2021
- Alco’s and a WWII Survivor March 23, 2021
- Comparing the 278A to the 567B Internally March 5, 2021
- Old Advertising X – Nordberg Radials February 6, 2021
- The Winton 201 at the Century of Progress January 2, 2021
- Tug Profiles – M. Moran December 8, 2020
- Delta Municipal Light & Power Part IV – Fairbanks-Morse 31A18 December 1, 2020
- Delta Municipal Light & Power Part III – Fairbanks-Morse 32E14 Engines November 26, 2020
- Delta Municipal Light & Power Part II – Fairbanks-Morse 33 Engines October 14, 2020
- Who IS Roots? And Why Does He Have a Blower Named After Him? October 3, 2020
- Delta Municipal Light & Power – Part I September 27, 2020
- Post Archives Volume II September 21, 2020
- A Home-built Locomotive – Southern San Luis Valley D-500 September 14, 2020
- Vintage Diesels In The Field September 6, 2020
- F-M Diesel-Electric Proposal August 16, 2020
- Don’t mess with our engine. August 10, 2020