Hacker News new | past | comments | ask | show | jobs | submit login
Howell Torpedo 1896 [video] (youtube.com)
72 points by gus_massa 16 days ago | hide | past | web | favorite | 26 comments

It's a really clever design, especially that gyroscopic 'impulse' mechanism, but it was never used in battle. The competing Whitehead torpedo, which had an onboard engine, was the one that navies adopted.


I get the distinct feeling that engineers could do more with a lot less back then. So much creativity in this design -- and essentially an analog implementation of P in a PID controller with the pulsed course corrections. It's awesome.

I've worked extensively with mechanical engineers on large scale projects. This level of creativity is very much alive today. We just don't get exposed to it as much here on HN.

As someone that loves the "How it's made" videos: Do you have some videos or explained graphics of your work that can post here? (As another submissions.)

One of the problems is that making these videos require a lot of time. This video is 6 minutes long, but I guess it was at least a week of work, perhaps more

I'm not sure how the flywheel speed is sensed, though. I don't think it was explained in the video, unless I missed something?

Also, I am clueless about imperial units (or rather, I make a conscious effort to not convert them...).

> I am clueless about imperial units

As am I. I had a laugh today when I stumbled across an interface in Siemens medical software that requires patient height and weight. I could enter ‘metric’ or ‘U.S.’ values.

> requires patient height and weight. I could enter ‘metric’ or ‘U.S.’ values

My scales has three settings, which I respectively think of as metric, US and British: kilos, pounds and stones+pounds.

This is likely because US imperial/standard isn’t the same as the old UK imperial. Height and weight were the same (though I don’t think the US uses stone) but volume is different; a US pint is not the same as a UK pint.

Might not be necessary as the intended range was a short 500 yards (~455 meters).

Well, the video does explain that the pitch of the propeller blades increases to compensate the flywheel slowing down: https://youtu.be/xTRBbFX7AxA?t=343

This just made me realize that hydraulic couplings could be (and are) used for continuously variable transmission systems :)

"Analog computing" is a great subject to get lost in on Wikipedia.

Some prime examples: Babbage's Analytical Engine, the Norden bombsight, the Mark 1 Fire Control Computer, etc.

That stair-step "gear" and pawl for pulsing the control vanes blew my mind.

That there are over a half dozen systems all functioning simultaneously further blew it. :-)

For those getting lost in steam era torpedos, you might also enjoy the earlier Brennan Torpedo. It was propelled and steered by hauling fine wires out of its stern with a land based steam engine.


My favorite part of that torpedo is the little flag. Because it didn't use steam, it didn't have any wake. So to be guided it had a little flag on a pole above the water.

The truly amazing thing about this is the flywheels double usage as stabilizer.

If the torpedo changed direction due to wind or waves the forces of the wheel cause the torpedo to bank.

This is then detected by a pendulum which sets the course straight again.

What an amazing work of engineering! I can't believe they figured all these out way back in 1896. How much of testing and iterations would have happened to make this work. I am super amazed!

Very clever way of control the rudders. Pure mechanical self navigation.

One of the world's first disposable [analogue] computers

How does this steer to the target or how did torpedos hold onto the target in the era of full mechanical torpedos?

I'm guessing it doesn't. It seems unguided. As I understood the video, all that machinery is there to ensure the torpedo swims in a straight line at a given depth, automatically countering any influence that tries to push it off course.

Thats why i was getting confused, I thought the same, but what would mean any influence, any current or perhaps waves? What else could stop a torpedo if not some target boat? Would a smaller boat or a swimmer be able to influence its trajectory if they were at the right time/place? Somehow doubt that. Im not very familiar with war/naval technology in general, just a basic idea of the principles and their physics

I'd say currents, detritus, and imperfections in torpedo shape. It's hard to make an inert thing move in a straight line in water, much like it's hard to make a paper airplane that flies straight. A control system like this, however, can maintain a flat trajectory.

It just travels in a straight line. Which is a pretty significant advancement over potentially turning around at the launching ship!

If I am not mistaken the gyroscope technique is also used in smaller satellites

I believe that is true as well.

Also, you reminded me, using the release of wound cables to de-spin a satellite. Very analog.


There is one on display at the free Naval museum near Paulsbo WA.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact