Work Day 5th August 2017

Six of the team were present for Saturday’s working session so a lot was achieved but not until grabbing the chance for a group photo!  [Your blogger was unfortunately not there due to family commitments].

Group Photo wm.jpg

LCP
Work continues on completing the painting of the LCP cabin’s rear wall. Holes were reamed and re tapped for the roof catches and primer applied to the places were using the paint roller is not practical.

T86
The majority of work is now focused on the T86 with several tasks being undertaken.

De rusting and priming the wheel arches inside the T86 cabin. The original flooring was removed some time ago as it was completely rotten due to water ingress and fortunately the steel sections of flooring had not corroded through, namely the wheel arches. Here you can see the stripped floor and the primed wheel arches.

T86 Wheel Arches Primed wm.jpg

And here the floor is being refurbished.

T86 Floor Refurbishment wm.jpg

Removing as many of the external fittings to the cabin as possible, as there is usually corrosion underneath them, specifically the reflectors (seen below). Not an easy task so an impact driver was required to remove the corroded in self tappers.

T86 Impact Driver Removing Reflectors wm.jpg

Rubbing down the end wall of the cabin. Unlike the LCP the T86 cabin is in much better condition paint wise so no going back to bare metal much to everyone’s relief – especially those working on it (live action below!).

Working on the T86 wm.jpg

Simulator

The simulator was run up for its usual weekly check for us to discover it has a fault. It boots to the initial Bloodhound display screens the console lights flash (usual start up sequencing) but not all together as they should and the simulator then freezes. No time on Saturday to pursue the fault but the suspicion is a missing input to the Digital Input box so the Argus hangs waiting for it ……… we think!

Off Site – Pedestal Gear Box Refurbishment
Refurbishing the Pedestal Gear box so the pedestal can be lowered. The gearbox has been dismantled and bearings removed, the majority of which are U/S due to corrosion again caused by  water ingress. Here are the bearings being removed using a puller.

Pedestal Gearbox pulling bearings.jpg

And here are the culprits:

Pedestal Gearbox  bearings.jpg

The bearings are imperial sizes but fortunately still available so a new set has been ordered.

Off Site – Test Rig Development for I/O Cards
The test rig for checking Digital Output cards is progressing well using an Arduino to set the required addressing and data required to test the card. The digital outputs are now controllable so the next stage is to attach LED’s to the switched outputs that correspond to the LED indications on the front of the card. The test rig which is the basis for eventually testing all I/O cards for the LCP MK2 consists of:

  • Laptop for programming the Arduino and displaying waveforms/logic levels from a PicoScope (an excellent bit of kit).
  • An Arduino
  • A locally designed Serial PeriBus generator which has six lines that connect to the I/O box with a 20 way ribbon cable
  • A salvaged Digital Output box to replicate the Digital I/O box in use in the LCP. The Digital I/O box containing a serial to parallel PeriBus converter card , a parallel PeriBus Termination card and the Digital I/O card under test.

Here is the rig in its development stage of build driving a Digital I/O card.

Simulator Driving Digital I-O Card.jpg

 

 

Work Day 29 July 2017

 

LCP Cabin
A long outstanding job was completed on the LCP cabin, the rubbing down of the cabin rear wall ready for painting. Next Saturday the rear of the cabin will be primed before a final top coat is applied.

Type 86 Radar
We need to ‘hand crank’ the aerial pedestal retraction mechanism to check the pedestal seal is OK with the underside of the cabin roof. We will also need to fully retract the pedestal to move the T86 at some time in the future. Problem, the mechanism is seized. The pedestal motor and the gear box that drives the chain for winding down the pedestal has been exposed to the elements and hence corrosion the likely cause. It is not required that the motor ever works again but the gearbox must be usable.

 

T86 Pedestal Motor.jpg

The motor was removed and the next check revealed the gear box (below) was seized (probably the motor as well). It was a struggle to split the gear box which confirmed the damage done by water ingress.

T86 Pedestal Motor Gear Box.jpg

When the T86 was recovered the pedestal cover above the retraction motor was not sealed so water simply ran down the pedestal, under the cover and on to the retraction motor. Work now starts on refurbishing the gear box so the pedestal can be hand cranked. It was interesting to find that two gear wheels were made of paxolin so when the pedestal retraction was driven by the motor and if the limit micro switches failed then the paxolin gears would strip preventing the motor trying the drive the pedestal though the radar roof or floor.

Work has also started on preparing the aerial system for a repaint so it is the usual removing of lose paint, corrosion treatment, filling and rubbing down before we get anywhere near a paint brush! Neil is hard at it again!

Prep on T86 Ae.jpg

Bloodhound 2 – Non-Coherent Jamming

A little while ago Pete Murray asked whether the Type 86 was vulnerable to non-coherent CW jamming. I reported in the Work Day blog for 22nd July that he tried it out on the simulator and it appears that the radar can be jammed by an non-coherent jamming signal at the skin echo frequency. As promised here is that more detailed report.If the jamming signal perfectly overlaps the skin echo, the radar will not acquire the jamming signal (as it fails the coherency check) and it cannot ‘see’ the skin echo.

The radar was able to acquire on a side band if the target was set-up to generate them.

Radar Display Non Coherent Jammer wm.jpg

Pete concludes that intelligent, autonomous jammers of the mid-eighties would have been able to jam the Type 86 radar using fairly simple non-coherent jamming techniques.

The Swiss modified their T87 radars to remove this vulnerability; it’s surprising the UK didn’t do likewise with the T86.

Here’s the theory:

  • Each type 86 had a unique ID modulation frequency. When a target is first detected, a coherency check is carried out on the received modulation. If the check fails, because the modulation frequency is wrong or absent, the target is rejected.
  • If a target is able to re-transmit a boosted copy of the skin echo without the ID modulation, the radar will not see (AGC out) the true return that carries the ID modulation.
  • If the jamming signal differs from the skin echo by any amount, the radar will see the skin echo.
  • Deviation of the ID FM widens with decreasing range, which might help the radar see the skin echo as range decreases.

The technology to analyse the radar signal and transmit a perfectly tailored jamming signal probably wasn’t available until the mid 80’s.

Pete thinks there must have been some sort of work around that allowed non-coherent jammers to be tracked, but he has not been able to find it. Otherwise, it seems odd to build this type of jamming into the RAF simulator.

Mach 1 in its own length … no way!

Brian Blestowe set out to debunk the urban myth that a Mk2 Bloodhound missile reached Mach 1 (speed of sound) by the time it had cleared the launcher; he was successful!

The Mk 2 missile was only stressed to 35g in the longitudinal axis so Zero to Mach one in its own length (around 1700g) was a bit of Newton BS. Actual acceleration at launch was 19 to 21g dependent on the ambient temperature of the boost rocket motor propellant charge. Actual specs for the Gosling XV motor were 23,000 lbs for 3.8 seconds at -25°C to 31,500 lbs for 2.8 seconds at +40°C. The only firing that I do have information about in the UK was the Singapore round fired in 1980. It was fired at a sea level temperature of 12.6°C and an atmosphere pressure of 1007.3 millibars.

Boost motor separation was at 3.803 seconds after first movement at which point the missile was 1332.6 m (4536.1 ft) from the launcher in ground distance and 795.0 m (2608.3 ft) above the launcher at a speed of 701.0 m/sec (2299.9 fps) (1568.09 MPH, 1362.63 Knots, Mach 2.04373). At that point the missile angle to the ground was 29.060° and it had rolled to starboard 15° off the line of fire (which was 328°), which wasn’t a
surprise seeing that the Wind velocity was 5.182 m/s (11.59 MPH) from a bearing of 250° (almost side on to the line of fire). For hotter (and colder) temperatures, the Swedes fired one at -25° C from Vidsel, while all of the hotter ones would have been from Woomera.

The problems getting figures from those firings is most of them were done with the original Gosling IV motor planned for BH2, which only produced 22,000 lbs at -25°C and because of that was not capable of accelerating the missile to a speed where the ramjets could produce enough thrust to accelerate the missile post boost separation on a really cold day (which was bad news for sales to Sweden and Switzerland!).

———————————

Brian is a BMPG member and the volunteer at Aerospace Bristol who has redied the Bloodhound Mk2 missile for exhibition. This is the only fully operational example that is complete; the only missile in UK to have fuze aerials and in every respect is the National Standard.

More about Bloodhound Homing

Pete Murray gave an expansion as to how the homing methed was achieved.

A very rough outline (from memory):
  • A gyro mounted on the dish allows its stabilisation via the dish servo (decoupling it from missile motion).
  • A spinning antenna offset from the spin axis creates a conical scan pattern. The phase and amplitude of the received signal reveal the dish pointing error (direction and magnitude of target misalignment from boresight). The dish servo steers the dish, via precession of the gyro, to null the dish pointing error.
  • The output of the dish servo represents the sight line rate in pitch and yaw. Zero sight line rate implies that the missile will intercept the target. Thus, to achieve an interception, the missile flight path must be controlled to null the sight line rates.
  • The pitch and yaw sight line rates are resolved into pitch and roll errors for the wing control servo.
    The roll channel moves the wings differentially to control to the missile roll rate and null the roll error, which brings all the error into the pitch plane. The Pitch channel moves the wings together to control missile flight path, such that an acceleration (around a curved path) proportional to the error is achieved. How hard the missile manoeuvres for a given error is controlled by a navigational constant
Other Notes (again, from memory):
  • The missile was monostable in the pitch plane (it would only pitch only one direction).
  • The missile had a weathercocking motion in flight; hence, two accelerometers displaced as far apart as possible were used in the pitch feedback loop.
  • A gyro provided roll rate feedback.
  • The navigational constant was halved for engagements against receding targets.
  • Gravity created a one G bias on the accelerometer output, which had to be compensated for to prevent the missile coming to earth during low altitude engagements.
  • The missile had characteristic body twisting and bending moments, which had to be filtered from the roll and pitch control channels.
  • Sightline rates were heavily filtered until switched out via the command link.
  • Navigation was in bearing only until climb cruise was switched out via the command link.

The Bloodhound Homing Algorithm

Back in May Peter Wolstenholme kindly reminded the BMPG Group about the Bloodhound homing algorithm.

The method of navigating to the correct interception point is called “proportional navigation” and relies on measurement of sight line spin (sls). That is, if the missile’s dish is pointing in a fixed direction in space, and continues to do that, then the missile will intercept the target. Any need to precess the dish gyros is caused by an error in the navigation path, so the missile needs to change direction so as to servo the sls to zero. The course change is set to a few times the sls.  An interesting control problem, as the computed Miss distance is equal to the sight line spin divided by V t^2.  V is the missile velocity and t^2 is the square of the time to go. So the gain of the servo increases rapidly as one approaches the target. Of course, at interception, if not a direct hit, the s.l.s. becomes larger than the dish servo can track, but by then one is within fuze range.

Peter added, “I spent many hours running end-course simulations for Bloodhound 1 to fix the optimal parameters, around 1957. End-course meant the last 5 seconds. I was very impressed by the agility I observed of twist-and-steer, conferred by the low rotational moment of inertia as compared with  missiles which needed to pitch from rear-mounted control fins”.