Hallade Method! Easily Finding the Radius of a Railway Curve With a Simple Piece of String

[@stevesellers-wilkinson7376]

Can't believe I've only just discovered this channel... particularly considering how much railway stuff I watch. I also work for Network Rail and this guy explains stuff really well. In my view, he should have stayed on as a teacher... so much better than the e-Learning (online courses) we have to do!

[@wernerviehhauser94]

I'd prefer to use lowercase letters for lines, since I can be sure that using C for chord length AND to denote a point in the plane would confuse the hell out if most if my students...

[@snarkybuttcrack]

hooray the last impediment to my global railway empire is cleared thanx to you kind sir

[@thepwayengineer]

😂😂😂

[@aimdrummer]

If you want to get the feel of how Hallade works, forget excel to start with, do it the old fashioned way with a pencil and an eraser (good luck finding proper Hallade paper though). See how the slues build up as you do the maths. To be fair, I sat in a classroom for two weeks being taught Hallade by (sadly no longer with us) Howard Finch and it is a very big subject to fit in to a short KZbin video but a good intro to the principle would be the old-school Chargehand's "averaging" method for removing localised alignment faults. Take three versine readings with the middle one being on the fault, work out what the average versine is of the three readings and slue the track to that versine on all three. It's not going to give you perfect alignment but for a localised fault it will improve things a bit.

[@thepwayengineer]

Hallade does seem to be a very much dying art. I struggled to find much material on it apart from speaking to people at work who are well versed in it. That does sound like a good place to start. Thank you

[@petes6814]

I actually enjoyed doing hallade schemes, excel definitely made it easier though. Still have some hallade sheets though.

[@reidflemingworldstoughestm1394]

Half the chord length squared, divided by the versine, then add the versine, and divide the whole thing by 2.

[@jimwinchester339]

Thanks. Very clear.

[@thepwayengineer]

Glad it was helpful!

[@aliendon73]

I always found the easiest way of calculating the radius of a curve was, by using a 10m string, measuring at half points (5m) to get the versine. Taking all the versine readings along the curvature. Then calculating the mean average, by adding all the versine readings and dividing by the amount of measurements taken, or using the most recurring versine reading. Then to work out the curvature using a simple calculation (via a caculator) 10m string = 12,500 divided by the versine. 20m string 50,000 divided by the versine. 5m string 3,125 divided by the versine. Using the scientific formula you use can confuse a lot of people. 20m = 50,000/V 10m = 12,500/V 5m = 3,125/V

[@paulkurilecz4209]

An excellent explanation

[@thepwayengineer]

Thank you!

[@randyshoquist7726]

Interesting. When I was a speed skater I worked out a method of measuring the blade's radius (rocker) using a straightedge and thickness gages. As I recall, the formula I came up with was R = C^2 / 8V + V^2 / 2. And when C >> V the second term is negligible, so, the same as your formula.

[@thepwayengineer]

Interesting! Never thought of speed skating as related in anyway!

[@randyshoquist7726]

@@thepwayengineer Analytic geometry. The great unifier.😁

[@mikefochtman7164]

When speaking of the radius, is it convention to use the outside rail vs the inside rail? Interesting video, thanks.

[@ojas3464]

👍After projecting a Sphere on to its Great Circle, Spherometer uses the same Formula to obtain the Sagitta, the analog of Versin

[@onlooker251]

I have been searching different sites for information of this type for some time to assist in the design and maintenance of a miniature railway which I support as a volunteer. Thank you for sharing this information. I have also subscribed to your channel. 👍

[@thepwayengineer]

Glad I could help!

[@andrewcharnock3]

Love your videos. But could you do a video on MSP measured shovel packing as this is a great permanent repair but is becoming a dying art. Would make a great education video thanks

[@neuralwarp]

C was a point on the diameter. How does C have a length?

[@aliuyar6365]

thnks

[@eric_on_youtube]

Will the radius ever vary along the curve -- increasing or decreasing?

[@thepwayengineer]

If the radius does change this is most likely the transition between two curves, this layout is known as a compound curve. Check out this video for more: kzbin.info/www/bejne/a5_RomWPprSlpLc

[@christianbarnay2499]

If you directly connect a constant radius arc to a straight line, there is a brutal lateral force applied on the train at the connection point. At slow speeds it's just an inconvenience to the passengers.But at higher speeds it can derail the train. To avoid this and allow high speeds in curves you need smooth transition curves with the radius changing progressively for the lateral force to build up and down progressively. The ideal form for those transition curves is a Euler spiral.

[@ronniel5941]

⁠​⁠@@christianbarnay2499I used to design trunk roads and motorways with similar principles to railway track design. These occasionally incorporated compound curves. One at A9 Dunkeld Bypass, changes radius five times, but that was unusual.

[@christianbarnay2499]

@@ronniel5941 Yes there is usually very little constraint in road curve design since the steering is handled by drivers. Having a straight line directly connected to an arc of circle with no transition is a frequent occurrence. Sometimes roads are even deliberately designed with hard transitions to force cars to slow down. But roads that are designed for high speed tend to copy some principles from rail tracks with wider tolerances since the drivers still have some control with their steering wheel.