The Northrop Bantam X4

This is a contribution from Giuseppe Ghisleri which originally appeared in RCM&E.

I live in Cremona, a little town in northern Italy, some 60 miles south-east of Milan. Cremona lies in the heart of the Padus valley, which is an absolutely flat piece of earth. Nearest alpine hills, to the north, dist some 30 miles. To the south the appennines, dist 20 miles. This not withstanding, I fly slope models with great pleasure, mostly in a site called Grone, in the Alps ,near Bergamo.

At Grone we fly scale models in the four meters class, Discus, DG 600, DG 800, and ASW 24 are a common sight. One guy builds giant models, in most cases they are 1:2 scale, which he launches by himself. These models weigh some 20 Kg or more, depending on the type reproduced. As the place is often overflown by full scale gliders coming from a nearby airfield, sometime one wonders if he is watching a model or at a man carrying plane.

The kind of flying in this environment is almost exclusively thermal and I have to say that I enjoy very much flying over the valley, chasing for thermals, and circling inside the one found, gaining height and than diving for aerobatics, or for a simple fly-by along the hillside.

Neverthless, being a modeller for more than 40 years, I am now 51, I have tried and enjoyed almost every kind of flying models, and when I see something new to me, I am almost every time attracted. To make a long story short, I once saw in an English magazine, some PSS models and fell in love with this category.

In Italy we do not have the kind of slope usually found in Atlantic facing countries. Suitable hills with lifting winds are a rare thing, but I was so involved that I can not resist and built an MB339 PSS model.

To fly this model I had to travel to the hills facing the gulf of Genoa, some 100 miles far from home. While I was lucky enough to find the weather ideal for flying, the right wind direction, the right wind force and a warm and sunny day that gave me a taste of what PSS flying is all about, I reasoned that, in my life, maybe I'll never find these favourable conditions all together again.

So I started thinking of a model which I could fly with a reasonable frequency from my usual flying site: Grone.

As I said previously, at Grone we fly in thermals, as is usually done in the Alps. But this site has a peculiarity : Some 200 meters to the right of the launching site, the hill falls abruptly for 900 meters to the bottom of the valley. This precipice exposes to the sun some naked rocks. You can easily imagine what this means. Almost invariably you can find in the afternoon, very strong and very large thermals.

So what I had to build was a model which can fly lift and thermals. A large, high aspect ratio airplane should have been the logical choice, say an U2, a B52 , just to name a few readily known. But I am sure that you know that a modeler seldom makes a reasonable choice. We follow our occasional love, when the time to build something has come.

At that time I was in love with the Northrop X4 Bantam, which I first saw in a book describing a completely different airplane: the F89 Scorpion. So reasoning that a flying wing is fast to build and should develop low drag, I decided what kind of wing profile to use and started.


The wing chord at the root is more than twice that at the tip. The cutting of the wing cores, is therefore a tough job. Having ruined two sets, I reverted to a friend of mine that owns a computer driven cutting machine. After some time, I got two perfectly cut cores, that required only sheeting. Root profile is an E224 ,tip profile an E230. According to Mr Eppler a wing with a sweep of 1 1/2 the mean aerodynamic chord, employing an E220 at the root and an E230 at the tip,is aerodynamically stable and does not need any tip washout. While not having the right sweepback, I used E224 as the root profile. This should gave an equally stable wing with less sweepback.

Back to construction.
Sheet the cores with 1/16" balsa or with obechi veneer. Cut elevons and face leading and trailing edges with 3/16" balsa. Note that elevons are hinged as is usually done on powered models, rather than on a soarer. This is done to use torque rods to drive them. Cut servo bays and drill in the foam a hole running from the bay to the elevon. Bend a 3 mm O.D. piano wire to form the torque rod.
Face with 1/16" ply, both servo bay and wing . These behave as torque rod bearings. Glue together wing halves and band with glass as done on powered multis. Insert torque rods. Mount a wheel collar and fix it in place on the torque rod with a long screw. Block in place the screw with cyano and then cut out screw head. You now have a complete torque rod.

Glue and shape wing leading edges and wing tips. Wing fences are cut from 1/16"ply. Complete wing by glueing in position hardwood dowels for wing holddown screws.

The fuse is made with blue foam and glass. This method is very similar to the one called "lost mould". It differs only for not having to dissolve the mould. I prefer to leave an hollowed core inside. This way you are not plagued with the mess of dissolved foam and you have a lighter construction having to use less glass.

I will not explain the method, but will outline some points that maybe can be easily overlooked.
It is easier to build the mould using two blocks of foam rather than one. The joining surfaces will give an accurate longitudinal axis. Hold together right and left blocks with double sided tape during shaping.
With this type of model it is difficult to have an accurately cut wing seating, so I thought to use the wing cushions for this job. More accurate than this is not possible.

Simply cut from fuselage halves two blocks having a rectangular shape in side view and the sweepback of leading and trailing edges on top view.
The heigth of the cut depends on that of the blocks used for the wing cores.
There is no real former in the fuselage, other than the one placed at the engine's exhausts. This serves as a template for shaping and gives an accurate planform to glue exhaust cones. Other pieces of balsa at the leading and trailing edges of the wing cushion have the same purpose. I usually fix wing on fuselage using wood self tapping screws. I find this method very easy to make and absolutely safe. The fin is removable to make transport easier. It remains in place during the flight without any type of fixing, only by drag between rods and tubes. Nose probe and pitot tube on the fin are both removable for obvious safety reason. All in all is an easy model to build.


The launch is difficult because of the almost flat bottom, so you need a friend to fly. Once in the air however, the X-4 is an absolutely stable model, very easy to fly, a little oversensitive in the pitch mode. If you own a computerized Tx, it should be possible to set commands so to have more servo displacement on the aileron mode than on the elevator mode.
This type of model can not stall. While you feed in more elevator, it will rotate upward its nose, losing speed until it will start to fall almost perpendicular to the ground like a dead leaf. This is very useful when flying from restricted area sites.

If you remember, I said that this model was conceived to fly in slope wind and strong thermals. Now I can say that I have got my goal. At Grone I fly positioning myself near the precipice, while a friend launches the model from the usual site, some 50 meters higher and 100 meters far away. I immediately direct the X-4 to the precipice. If the conditions are good you can see immediately when it reaches the edge as it jumps up like if you had fed in full power. The X-4 flies circles in thermals at high speed and goes up very quickly. Gains speed while diving and executes large loops with ease.
I do not have yet tried inverted flight. I have a video tape showing the original aircraft along with the De Havilland "Swallow". I have to say that the flying of the model matches that of the prototypes.
Once the precipice lost its thermal activity suddenly, leaving the X-4 in trouble. I had to let it go down in the valley, searching for ascending air. While it was slowly losing height, I thought that my model was lost, along with servos and Rx, because there is no way to find a model in the thick wood that covers the almost vertical slope.
Suddenly I realized that the Bantam was no longer sinking, so I started circling in the light thermal found. Slowly, almost too slowly for my nerves, the X-4 started gaining height as the thermals reinforced. To make it short, I landed the model at my foot gaining a great cheer from fellow modelers watching the performance.
Later I brought the model on a lift site, but the conditions were very marginal. I tried to fly and realized that, for being a PSS, it is a real floater. I hope to find sooner or later, a good day to fly on strong lift. I am sure that the X-4 will give me much pleasure.

With the C.G. shown on plan, the model needs up elevons to fly level, 3-4 mm at the root trailing edge. Of course this condition is of detriment for the overall efficiency. I have tried to move C.G. back just a little, say 2-3 mm, and the result was as expected. Less up trim, but more sensitive in the pitch mode. Having such a trouble to find proper conditions, I am not very prone to waste flying time just to investigate so I fly the X-4 as it is and enjoy it very much! Probably it would benefit from a change of root profile, from E224 to E226,but this way the wing will lose some of its lifting ability.


In June 1946 ,Northrop received from the U.S. Air Force an order for two X-4 airplanes. This airplanes's task was to investigate the manouverabilty that tailless airplanes would have in the region of transonic flight. Its design was influenced from the German war-time Me 163 Komet and from the De Havilland D.H.108 Swallow.
It reached mach 0.94, but at this speed it had serious stability problems. The wing was very stiff with a loading factor of 8G. The aircraft was very manouverable and could do loops and rolls and other aerobatics with ease. It was powered from two Westinghouse jet engines, each rated at 725 Kg of thrust.
The cockpit canopy had no stiffener and it was the first X-plane to have a jettisonable seat. One still exists, restored in an American museum.

E-mail Andy MacDonald