Cruisers Forum
 


Reply
 
Thread Tools Search this Thread Rate Thread Display Modes
Old 24-10-2017, 02:26   #586
Moderator
 
Dockhead's Avatar

Cruisers Forum Supporter

Join Date: Mar 2009
Location: Denmark (Winter), Helsinki (Summer); Cruising the Baltic Sea this year!
Boat: Cutter-Rigged Moody 54
Posts: 28,478
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
And you were wrong to propose the crossing maintaining a 1/4 nm CPA from the stern of the lead ship, at least in my humble opinion.
OK, now we're talking about "humble opinions". This is good It's a really interesting subject -- let's try to discuss it rationally.

The key here is the phrase "maintain" a 1/4 mile CPA. You can't maintain it. We need a simulator here so that you can see what this crossing looks like as it unfolds. By the time you get even with his bow, you are long since committed to the crossing. You have to set it up much earlier than this. So you don't "maintain" a 1/4 mile CPA, you set it up from two or three miles away. You can't just sail up to his side and jink around -- by the time you get even with his bow, it is too late for you to correct anything. A 1/4 mile CPA, or a couple of cables anyway, is the closest crossing you can set up to even get there and pass safely. As someone said, 180 feet is the same as 0.

And Lodesman is right -- of course a prudent mariner just wouldn't do this crossing at all, and certainly not without clear agreement by radio with both ships. If 1/4 mile becomes a half mile, which it can easily do, then you're screwed.
__________________
"Parce que je suis heureux en mer, et peut-ętre pour sauver mon ame. . . "
Dockhead is online now   Reply With Quote
Old 24-10-2017, 03:35   #587
Registered User

Join Date: Nov 2005
Location: Out there doin' it
Boat: 47' Olympic Adventure
Posts: 3,482
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
I believe, that if they would stand on for one to be 1/4 nm away from their port transom, they would likely also stand-on for one to be 180 ft from their port transom.

Agreed. Waiting out the convoy is an option, and definitely the safest without question. What we are considering is that if one decides to go, what is the safer approach, 1/4 nm from the lead boat port stern, or 180 ft. (I believe the latter, though I wouldn't necessarily do it, I would be more likely to just go somewhere else.)

Even worse if one maintained a minimum 1/4 nm CPA from the lead ship port transom, which is exactly my point.
I'm not certain any merchant watchkeeper would be happy with a 1/4 mile passing either, unless they've talked to you by VHF. Even then a lot of that depends on how well they can view the interaction. A typical vessel with the bridge towards the aft end of the ship is likely to have a better appreciation that you will pass clear aft, than one with the bridge forward.

This brings up a point - how are you measuring your pass? Just eyeballing a bearing change doesn't give the range, and certainly not the CPA. AIS is out the window, and I doubt most radars would give that sort of definition.

How close is too close for you? 120 ft? 80 ft? 0 ft?

I do agree that passing 1/4 mile astern of the first vessel is not very safe, but there is no way for you to pass 1 mile ahead of the second vessel. As I said the maximum passing from that ship would be about 3 c.
Lodesman is offline   Reply With Quote
Old 24-10-2017, 03:55   #588
Moderator
 
Dockhead's Avatar

Cruisers Forum Supporter

Join Date: Mar 2009
Location: Denmark (Winter), Helsinki (Summer); Cruising the Baltic Sea this year!
Boat: Cutter-Rigged Moody 54
Posts: 28,478
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by Lodesman View Post
. . . I do agree that passing 1/4 mile astern of the first vessel is not very safe, but there is no way for you to pass 1 mile ahead of the second vessel. As I said the maximum passing from that ship would be about 3 c.
Yes, and I failed to fully visualize this myself. I've done a lot of threading through lines of ships in real life. But I am usually making 8 or 9 knots, and they are usually more like 2 miles apart, so it's MUCH safer than this hypothetical. So if you get reasonably close to the ship ahead you can pass reasonably comfortably ahead of the second one, despite the real life variations like the ships aren't exactly in line, the crossing is not exactly perpendicular, etc.

As the difference in speed increases, the Relative Motion Line gets steeper and steeper, and this kind of crossing rapidly becomes more and more dangerous. The hypothetical case COULD be done I think, but it's really dangerous. You'd have to nail the crossing close behind and you would absolutely have to agree the whole thing with both vessels, as you suggested.
__________________
"Parce que je suis heureux en mer, et peut-ętre pour sauver mon ame. . . "
Dockhead is online now   Reply With Quote
Old 24-10-2017, 04:21   #589
Moderator
 
Dockhead's Avatar

Cruisers Forum Supporter

Join Date: Mar 2009
Location: Denmark (Winter), Helsinki (Summer); Cruising the Baltic Sea this year!
Boat: Cutter-Rigged Moody 54
Posts: 28,478
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
Nope, not my plan at all.

My proposed solution was to sail up on a perpendicular course to ship, coming within 180 ft of the port transom. (Originally incorrectly posted as 180 ft aft of the ships port transom. My bad, corrected in earlier post.)

Your proposed solution was to sail up on a perpendicular course coming within 1/4 nm of the ship. That puts you needlessly close to the oncoming ship, crossing it's bow.
And it was already explained why you simply can't do it, in anything like a seamanlike manner.

I see that you just don't visualize the relative motion of the vessels in this scenario -- this is clear from your surprise that he will be almost on your beam as you approach him. The crossing from one mile looks like this:

Click image for larger version

Name:	24_10_2017 11_23 Office Lens (1) (002).jpg
Views:	39
Size:	424.3 KB
ID:	158235

The point you are going to reach (IF both vessels continue as if on rails, and IF the courses are exactly perpendicular) will always be 14 degrees ahead of your beam. At one mile, the bearing of his bow and the bearing of your desired point 180 feet off his quarter will differ by only 4 degrees (if he's 400 meters long as Lodesman drew).

That's too close to set up reliably, and once you get close, you can't correct any more -- you can't move far enough, in the time you have.

So the two or three cables suggested, was suggested because this is just about the ragged edge of what you might manage if you are really good, without creating a big risk of getting under his bows as you approach.

In this drawing, the Relative Motion Speed is 20.6 knots, so this picture is what you see 174 seconds before, err, impact. At one mile, one degree is 300 feet. So if his course varies by one degree to port (or if you mis-estimated his course by one degree), your 180 foot CPA is GONE -- the Relative Motion Line will cross you. Since you cannot estimate his course with that degree of accuracy by any means, it means you cannot set this up.
__________________
"Parce que je suis heureux en mer, et peut-ętre pour sauver mon ame. . . "
Dockhead is online now   Reply With Quote
Old 24-10-2017, 04:53   #590
Moderator
 
Dockhead's Avatar

Cruisers Forum Supporter

Join Date: Mar 2009
Location: Denmark (Winter), Helsinki (Summer); Cruising the Baltic Sea this year!
Boat: Cutter-Rigged Moody 54
Posts: 28,478
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Drifting the point a little -- to Lodesman's question about HOW you would do it. To shave it as close as possible -- to get as close as you can get behind him, consistent with reasonable seamanship -- you must have the most accurate data possible, and you need a good plot. For all of the failings of AIS which we have discussed, AIS is going to be the best we have. It just must not be used to try to pass closer than the limitations of the system allow you to know. About a ship's length, I think, is about as close as you can have reasonable confidence, and that should take care of variations in antenna placement.

Then, as you get close, check the range with radar and verify the solution. Small boat radars are carp for bearing, but some of them are very good indeed for range -- my 4G set has been checked carefully. As you approach, radar will show the range to the closest part of the ship as the errors of AIS get more and more significant.

With a 4:1 speed difference and such a steep Relative Motion Line, it's a bit like landing a plane. You can't just fly close to the ground and then spontaneously push the stick when you glimpse the runway -- you have to set up a glide path from a reasonable approach altitude.

Certainly, this is NOT the time to "get your head out of the electronics" -- on the contrary, you need to have a crystal clear tactical picture based on data which your eyes cannot perceive. In hairy traffic situations, I have an experienced crew on the helm with his eyes open (and with a HBC, and he has his own radar screen), and I am at the nav table with the electronics and radios and plotting sheets and notebook. We compare perceptions by intercom.
__________________
"Parce que je suis heureux en mer, et peut-ętre pour sauver mon ame. . . "
Dockhead is online now   Reply With Quote
Old 24-10-2017, 07:26   #591
cruiser

Join Date: Jan 2012
Location: Lake Ontario
Boat: Ontario 38 / Douglas 32 Mk II
Posts: 3,250
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by Dockhead View Post

The key here is the phrase "maintain" a 1/4 mile CPA. You can't maintain it.
SAY WHAT!?!??!!?

Incorrect.

If the CPA is 1/4 nm, it is 1/4 nm for the entire crossing, else it is not a CPA. The distance off the ship varies constantly during the crossing, with 1/4 nm being the "Closest Point of Approach" (CPA), and this is maintained during the entire crossing, for every crossing, unless the skipper doubles back to get closer.

If you plan a 1/4 nm CPA, and your maneuvering actions keep that value, you have "maintained" a 1/4 nm CPA from the vessel.

Quote:
A 1/4 mile CPA, or a couple of cables anyway, is the closest crossing you can set up to even get there and pass safely.
We obviously disagree on this.

Without any need for a simulator lets look at the geometry shall we?

Scenario Recap:

There is a convoy of 50 commercial ships in open sea heading 0 degrees north in single file. They are all travelling at 20 knots and are 1-1/4 miles apart, bow to stern.

You are motoring in a sailboat 90 degrees east (give way) at 5 knots and need to cross the course of the convoy. Winds are steady, seas are calm, visibility is 20 miles. How would you do it?

The ships are 600 ft long x 100 ft wide, and the sailboat is 40 ft long x 15 ft wide.


Assuming the skipper commits to crossing rather than waiting for the entire convoy to pass (best plan in my opinion)...

Option A) Ramblin Rod proposal - 180 ft eyeball CPA.

From some distance out, the skipper sets up a crossing course 90 degrees (perpendicular) to the ships' path.

Before and after the point where the sailboat is 90 degrees and 180 ft from the leads ships port stern corner, the distance off is greater.

So lets say at ~ 8.25 nm off (distance between boats) the sailboat sets up the 90 degree crossing. The sailboat is 2 miles from the course of the ships port stern corner, and the ship is 8 miles out from the course of the sailboats bow.

Since in the real world there are variations in wind, current, and waves, there will be some (actually constant) course adjustments required of the sailboat, as there always is to hold any course.

So the sailboat plugs away, maintaining course to maintain a 180 ft CPA from a leading ships port stern quarter. This CPA is achieved 90 degrees off the vessels course.

The sailboat must travel, the 180 ft to the ships port stern corner course, and then another 100 feet to cross the stern of the ship with the bow, and another 40 feet to cross with the transom, for a total distance of 320 ft.

So this 320 ft of sailboat travel will take 320 ft / 6000 ft/nm = 0.05 nm / 5 knots = .01 hours or 0.64 minutes.

In that time, the sailboat transom will be 20 knots * .01 hours = 0.21 nm or 1280 ft aft of the lead ship and 1.0 nm (6000 ft) ahead of the following ship.

By the time (0.05 hrs or 3 minutes) the following ship passes the transom the sailboat, the distance between the two will be 0.26 nm (1555 ft).

Option B - Dockhead Proposal - 1/4 nm AIS CPA.

In this case, the crossing is set up so that the sailboats bow is 1/4nm away from the course of the ships port stern corner at 90 degrees.

The sailboat must now travel, the 1500 ft to the ships port stern corner course, and then another 100 feet to cross the stern of the ship with the bow, and another 40 feet to cross with the transom, for a total distance of 1640 ft.

So this 1640 ft of sailboat travel will take 1640 ft / 6000 ft/nm = 0.27 nm / 5 knots = .055 hours, or 3.28 minutes.

In that time, the sailboat transom will be 20 knots * .055 hours = 1.09 nm or 6560 ft aft of the lead ship and 0.16 nm (960 ft) ahead of the following ship.

The ship at this point will be 0.008 hrs (.48 minutes or 28.8 seconds) away.

By the time (0.008 hrs or 0.48 minutes, or 28.8 seconds) the following ship passes the transom the sailboat, the distance between the two will be 0.4 nm (240 ft).

So in summary, from being crushed by the following ship:

a) Ramblin Rods proposal clears the path of the following 20 knot ship by 1555 ft.

b) Dockheads proposal clears the path of the following 20 knot ship by 240 ft.

Remember this 240 ft is AIS CPA (as Dockhead chooses to prefer electronics over eyeballs).

So with the inaccuracies of AIS, there very well could be a big "BANG, GLUG, GLUG. GLUG".

Everyone is free to make their own decisions (and live (or not) with the consequences) but I would far sooner be a 180 ft from the lead ship and 1555 ft from the following ship, than 1500 ft from the lead ship and 240 ft from the following ship.

And this my friends is why one needs to get their heads out of the electronics, and get eyeballs on the situation when maneuvering in close quarters.

Case closed.
ramblinrod is offline   Reply With Quote
Old 24-10-2017, 08:21   #592
Registered User
 
Juho's Avatar

Join Date: Apr 2015
Location: Finland
Boat: Nauticat 32
Posts: 973
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by Dockhead View Post
Drifting the point a little -- to Lodesman's question about HOW you would do it. To shave it as close as possible -- to get as close as you can get behind him, consistent with reasonable seamanship -- you must have the most accurate data possible, and you need a good plot.
The title of this thread talks about cones of uncertainty. I don't want to sail in front of fast and large ships. In my mind I have a danger zone there. It resembles a cone. Let's say that I don't want to be within a 25° cone in front of a fast and large ship when it is 5 miles away (my danger zone is relatively wide at 5 miles).

If I get the location of the ship from AIS, I may need to add the length of the ship to the distance, since I'm not sure where the GPS antenna is. But let's say the rear end of the ship is now max 5 miles away (from the point where it would pass my current location).

At 5 miles I should be 2.1 NM (= 5 NM * sin(25°)) away from the estimated track of the ship. My electronic devices tell me that I'm 2.5 miles away. That's good enough for me, and includes also some reserve for CPA and the unknown shape and small random movements of the ship.

The speed difference is 4:1, so my most efficient angle (deviation from 90°), if I want to cross the track of the ship as close to its transom as possible, is asin(1:4) = 14.478°. I note that it is smaller than 25°, and therefore I can't approach the ship too fast. I can now switch to my track crossing mode. I will turn the boat to a 90°-14.4° course (in relation to the true course of the ship).

My boat (its GPS antenna) will reach the track of the estimated middle of the transom of the ship (2.5 NM / tan(14.478°)) - 5 NM = 4.682 NM behind the transom.

I think this is something that could happen in real life, when I want to cross close to the stern of the ship.
Juho is offline   Reply With Quote
Old 24-10-2017, 08:26   #593
Moderator
 
Dockhead's Avatar

Cruisers Forum Supporter

Join Date: Mar 2009
Location: Denmark (Winter), Helsinki (Summer); Cruising the Baltic Sea this year!
Boat: Cutter-Rigged Moody 54
Posts: 28,478
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
SAY WHAT!?!??!!?

Incorrect.

If the CPA is 1/4 nm, it is 1/4 nm for the entire crossing, else it is not a CPA. The distance off the ship varies constantly during the crossing, with 1/4 nm being the "Closest Point of Approach" (CPA), and this is maintained during the entire crossing, for every crossing, unless the skipper doubles back to get closer.

If you plan a 1/4 nm CPA, and your maneuvering actions keep that value, you have "maintained" a 1/4 nm CPA from the vessel.



We obviously disagree on this.

Without any need for a simulator lets look at the geometry shall we?

Scenario Recap:

There is a convoy of 50 commercial ships in open sea heading 0 degrees north in single file. They are all travelling at 20 knots and are 1-1/4 miles apart, bow to stern.

You are motoring in a sailboat 90 degrees east (give way) at 5 knots and need to cross the course of the convoy. Winds are steady, seas are calm, visibility is 20 miles. How would you do it?

The ships are 600 ft long x 100 ft wide, and the sailboat is 40 ft long x 15 ft wide.


Assuming the skipper commits to crossing rather than waiting for the entire convoy to pass (best plan in my opinion)...

Option A) Ramblin Rod proposal - 180 ft eyeball CPA.

From some distance out, the skipper sets up a crossing course 90 degrees (perpendicular) to the ships' path.

Before and after the point where the sailboat is 90 degrees and 180 ft from the leads ships port stern corner, the distance off is greater.

So lets say at ~ 8.25 nm off (distance between boats) the sailboat sets up the 90 degree crossing. The sailboat is 2 miles from the course of the ships port stern corner, and the ship is 8 miles out from the course of the sailboats bow.

Since in the real world there are variations in wind, current, and waves, there will be some (actually constant) course adjustments required of the sailboat, as there always is to hold any course.

So the sailboat plugs away, maintaining course to maintain a 180 ft CPA from a leading ships port stern quarter. This CPA is achieved 90 degrees off the vessels course.

The sailboat must travel, the 180 ft to the ships port stern corner course, and then another 100 feet to cross the stern of the ship with the bow, and another 40 feet to cross with the transom, for a total distance of 320 ft.

So this 320 ft of sailboat travel will take 320 ft / 6000 ft/nm = 0.05 nm / 5 knots = .01 hours or 0.64 minutes.

In that time, the sailboat transom will be 20 knots * .01 hours = 0.21 nm or 1280 ft aft of the lead ship and 1.0 nm (6000 ft) ahead of the following ship.

By the time (0.05 hrs or 3 minutes) the following ship passes the transom the sailboat, the distance between the two will be 0.26 nm (1555 ft).

Option B - Dockhead Proposal - 1/4 nm AIS CPA.

In this case, the crossing is set up so that the sailboats bow is 1/4nm away from the course of the ships port stern corner at 90 degrees.

The sailboat must now travel, the 1500 ft to the ships port stern corner course, and then another 100 feet to cross the stern of the ship with the bow, and another 40 feet to cross with the transom, for a total distance of 1640 ft.

So this 1640 ft of sailboat travel will take 1640 ft / 6000 ft/nm = 0.27 nm / 5 knots = .055 hours, or 3.28 minutes.

In that time, the sailboat transom will be 20 knots * .055 hours = 1.09 nm or 6560 ft aft of the lead ship and 0.16 nm (960 ft) ahead of the following ship.

The ship at this point will be 0.008 hrs (.48 minutes or 28.8 seconds) away.

By the time (0.008 hrs or 0.48 minutes, or 28.8 seconds) the following ship passes the transom the sailboat, the distance between the two will be 0.4 nm (240 ft).

So in summary, from being crushed by the following ship:

a) Ramblin Rods proposal clears the path of the following 20 knot ship by 1555 ft.

b) Dockheads proposal clears the path of the following 20 knot ship by 240 ft.

Remember this 240 ft is AIS CPA (as Dockhead chooses to prefer electronics over eyeballs).

So with the inaccuracies of AIS, there very well could be a big "BANG, GLUG, GLUG. GLUG".

Everyone is free to make their own decisions (and live (or not) with the consequences) but I would far sooner be a 180 ft from the lead ship and 1555 ft from the following ship, than 1500 ft from the lead ship and 240 ft from the following ship.

And this my friends is why one needs to get their heads out of the electronics, and get eyeballs on the situation when maneuvering in close quarters.

Case closed.
Now we're talking about real geometry -- that's progress. Let's talk about this idea of "maintaining" your 180 foot CPA.

How do you plan to do this? How do you plan to even see it? That's the whole problem, which you still don't see.

Here you are one mile off:

Click image for larger version

Name:	24_10_2017 11_23 Office Lens (1) (002).jpg
Views:	50
Size:	424.3 KB
ID:	158241


You are 174 seconds before CPA, and the ship is less than two points off dead abeam. How do you propose to see the difference, by any means at all, between getting to where you need to be, and running up under his bows? Which is a difference of only 4 degrees (and less than that with a smaller ship). 180 feet in this situation is equal to 0 feet -- you can't see or measure it from far enough away, to take action. You might even be passing ahead of him, and you won't be able to see or measure it, and you only have 174 seconds. And that's REALLY dangerous -- if you discover at 5 cables off that you're actually passing slightly ahead, what do you do then? The first problem is that a 180 foot CPA is really a zero CPA, because you can't possibly measure the crossing to that degree of accuracy. So the approach itself is unsafe, unless you are aiming for a point far enough from his quarter that are you able to have some reasonable idea that you are actually passing behind him.

At some point you will start to be able to measure how you are actually crossing, and it becomes apparent first, with most exact data -- that is AIS. Your eyeball can well understand a crossing with similar speeds -- where the Line of Relative Motion is shallower and you have more aspect on him. But at 4:1 you can't tell anything until you are almost under his bows -- he seems to be running straight into your beam. Note also that we take it as given that the courses are exactly perpendicular, but in real life they never are -- if his real course is even 14 degrees different, then he will be heading exactly towards your beam. Ships also have LEEWAY, so if the wind is blowing, his COG may diverge from his heading by quite a bit, you can't see this at all.

ALL THAT is why you have to aim for a point which is far enough away from his quarter that you have at least a reasonable chance of getting there. You can try to correct when you get close, but you will have only seconds, and you are making only 5 knots. Your best chance of passing within a cable of his quarter is to head kamikaze-fashion into a collision situation, using the best data you can get (AIS and radar to verify actual range), and head off at the last moment to avoid getting under his bows, which will be hellishly difficult to judge and really dangerous. Then turn hard to run down his side, and try to judge when you can turn to pass behind him, without banging into his quarter -- really hard when your combined speeds are 25 knots (!).

This is totally unseamanlike, violating several COLREGS, but it's the ONLY way you can get within a cable of his quarter with a 4:1 speed differential.

The seamanlike way is to set up a CPA from a couple of miles out which is large enough to be able to measure -- large enough to be a meaningful, correctable number. Get onto that glide path early and hold your course while closely monitoring AIS and radar and making corrections when it's possible -- when you have meaningful data.

I say again, when he's running at you at 4 times your speed, less than two points off dead abeam, from a mile or two away, your eyeball tells you NOTHING. A couple of degrees plus or minus will put you under his bows, on the one hand, or give you a missed approach, on the other, where you end up too far away to get back to where you could pass safely in front of the next ship, and you cannot see the difference between 16 degrees and 12 degrees from dead abeam of you, with your bare eyes, nor do you even know within a point or even two, how close to perpendicular his course is to you. You cannot do this at all if you don't understand the limitations of what you can see and perceive.

What you propose to do is like trying to land plane by flying really close to the ground, and then slamming the stick forward at the moment you glimpse the runway. Thinking that by approaching from much closer to the ground, you can land in a shorter distance. It doesn't work like that. If you are trying to land on a really short runway, you need all the more to approach from the correct glide path, in order to see the spot you are going to land from further away, and have more control over getting there. You don't aim for a spot which is three feet from the fence -- you aim for a spot far enough from the fence that it is within what you can reasonably know, so that your corrections are small and not desperate and last minute.

Just the same in this situation -- if you set up from a couple miles away a crossing with a CPA which is within what you can meaningfully measure, then you actually have a chance to make corrections, and you actually have a chance of passing somewhere near the point you aimed at. Instead of wild maneuvers, done by eye, at the last second, which are just as likely to put you under the bows of the next ship as anything.


The biggest ship in the world, travelling at 20 knots, will be GONE and past you, in less than 30 seconds from the time his bow goes past you, to the time his transom goes past you -- 1/5 of a nautical mile of ship flying past as you bob around less than a cable away. You can only move maximum 300 feet in that amount of time, at 5 knots. See Lodesman's plot of this -- in order to pass 180 feet from his quarter, you have to be already LESS THAN A CABLE from his bow, then as his bow flashes past, you have to keep running at top speed directly at him, during the 30 seconds it takes for him to flash past. How little you visualize this is evident when you talk about "doubling back" -- you can't get there, from there. He's not anchored -- and when you get close, it looks like a bullet going by.
__________________
"Parce que je suis heureux en mer, et peut-ętre pour sauver mon ame. . . "
Dockhead is online now   Reply With Quote
Old 24-10-2017, 10:52   #594
Marine Service Provider

Join Date: May 2011
Location: PDX
Boat: Gulfstar 50
Posts: 1,831
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
SAY WHAT!?!??!!?

Incorrect.

If the CPA is 1/4 nm, it is 1/4 nm for the entire crossing, else it is not a CPA. The distance off the ship varies constantly during the crossing, with 1/4 nm being the "Closest Point of Approach" (CPA), and this is maintained during the entire crossing, for every crossing, unless the skipper doubles back to get closer.

If you plan a 1/4 nm CPA, and your maneuvering actions keep that value, you have "maintained" a 1/4 nm CPA from the vessel.

SNIP %<
Rod, Let's get your head out of the electronics.

First off as your sail boat yaws back and forth as you try to keep a constant heading the AIS is showing CPA all over the place. No real use there.

Second, Eyes are good at seeing movement but not so good at judging distances and angles. Once an object is more than (oh say) 40' away where the parallax between each eye is used for distance then the brain uses expected size to estimate distances. Very imprecise.

Thirdly, you said yourself that this is not a good crossing. All can plainly see that this is a contrived example and not real world. But you continue to promote your solution as a good one.

Fourthly, even if there were 50 ships in a row they would not likely be all in line, not likely be 1.25 nm apart (make that 3.75 minutes apart - not much time to react).

Let me toss in another thought..............................

The sailboat under power is "shall not impede" and the ships are stand on. As the sailboat gets close (say) within 1 nm of the ships path the ships decide that the sailboat presents a risk for collision (Rule 7). Specifically 7a and 7dii.

The ship which the sailboat is attempting to pass astern of is likely to attempt to stop, alter course to STB and coordinate collision avoidance with all the other following ships which are coming up in line.

Chaos rules supreme

Let me point you to Rule 8f

(f)
(i) A vessel which, by any of these rules, is required not to impede the
passage or safe passage of another vessel shall, when required by the
circumstances of the case, take early action to allow sufficient sea
room for the safe passage
of the other vessel.
(ii) A vessel required not to impede the passage or safe passage of
another vessel is not relieved of this obligation if approaching the other
vessel so as to involve risk of collision
and shall, when taking action,
have full regard to the action which may be required by the rules of this
part.
(iii) A vessel, the passage of which is not to be impeded remains fully
obliged to comply with the rules of this part when the two vessels are
approaching one another so as to involve risk of collision.
evm1024 is online now   Reply With Quote
Old 24-10-2017, 11:04   #595
cruiser

Join Date: Jan 2012
Location: Lake Ontario
Boat: Ontario 38 / Douglas 32 Mk II
Posts: 3,250
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by Juho View Post
The title of this thread talks about cones of uncertainty. I don't want to sail in front of fast and large ships. In my mind I have a danger zone there. It resembles a cone. Let's say that I don't want to be within a 25° cone in front of a fast and large ship when it is 5 miles away (my danger zone is relatively wide at 5 miles).

If I get the location of the ship from AIS, I may need to add the length of the ship to the distance, since I'm not sure where the GPS antenna is. But let's say the rear end of the ship is now max 5 miles away (from the point where it would pass my current location).

At 5 miles I should be 2.1 NM (= 5 NM * sin(25°)) away from the estimated track of the ship. My electronic devices tell me that I'm 2.5 miles away. That's good enough for me, and includes also some reserve for CPA and the unknown shape and small random movements of the ship.

The speed difference is 4:1, so my most efficient angle (deviation from 90°), if I want to cross the track of the ship as close to its transom as possible, is asin(1:4) = 14.478°. I note that it is smaller than 25°, and therefore I can't approach the ship too fast. I can now switch to my track crossing mode. I will turn the boat to a 90°-14.4° course (in relation to the true course of the ship).

My boat (its GPS antenna) will reach the track of the estimated middle of the transom of the ship (2.5 NM / tan(14.478°)) - 5 NM = 4.682 NM behind the transom.

I think this is something that could happen in real life, when I want to cross close to the stern of the ship.
Sensible.

I liken the safety zone around a vessel like a tear drop.

The closer to the transom, the nearer the "safe distance".

The further from the transom, the greater the "safe distance".

There is a minimum distance, as one needs to contend with variations in wind, waves, current, visibility, etc. The greater the possible error induced by the variables, the greater the safe distance.

I completely concur, for the given scenario, the further distance one can maintain from the following ship, the better, and if that means getting very close to the port transom corner of the lead ship, sobeit.

Will there be dirty shorts on the lead ship? Possibly.

Will the sailboat skipper be on high alert, maintaining a watchful eye on the lead ship during the entire crossing? Sure better.

Should he be dicking with AIS, Radar, GPS, and Marpa AFTER he is in close quarters with the ship and controlling the crossing by eyeball? Not on his life!
ramblinrod is offline   Reply With Quote
Old 24-10-2017, 11:23   #596
cruiser

Join Date: Jan 2012
Location: Lake Ontario
Boat: Ontario 38 / Douglas 32 Mk II
Posts: 3,250
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by Dockhead View Post
Now we're talking about real geometry -- that's progress. Let's talk about this idea of "maintaining" your 180 foot CPA.

How do you plan to do this? How do you plan to even see it? That's the whole problem, which you still don't see.

Here you are one mile off:

Attachment 158241


You are 174 seconds before CPA, and the ship is less than two points off dead abeam. How do you propose to see the difference, by any means at all, between getting to where you need to be, and running up under his bows? Which is a difference of only 4 degrees (and less than that with a smaller ship). 180 feet in this situation is equal to 0 feet -- you can't see or measure it from far enough away, to take action. You might even be passing ahead of him, and you won't be able to see or measure it, and you only have 174 seconds. And that's REALLY dangerous -- if you discover at 5 cables off that you're actually passing slightly ahead, what do you do then? The first problem is that a 180 foot CPA is really a zero CPA, because you can't possibly measure the crossing to that degree of accuracy. So the approach itself is unsafe, unless you are aiming for a point far enough from his quarter that are you able to have some reasonable idea that you are actually passing behind him.

At some point you will start to be able to measure how you are actually crossing, and it becomes apparent first, with most exact data -- that is AIS. Your eyeball can well understand a crossing with similar speeds -- where the Line of Relative Motion is shallower and you have more aspect on him. But at 4:1 you can't tell anything until you are almost under his bows -- he seems to be running straight into your beam. Note also that we take it as given that the courses are exactly perpendicular, but in real life they never are -- if his real course is even 14 degrees different, then he will be heading exactly towards your beam. Ships also have LEEWAY, so if the wind is blowing, his COG may diverge from his heading by quite a bit, you can't see this at all.

ALL THAT is why you have to aim for a point which is far enough away from his quarter that you have at least a reasonable chance of getting there. You can try to correct when you get close, but you will have only seconds, and you are making only 5 knots. Your best chance of passing within a cable of his quarter is to head kamikaze-fashion into a collision situation, using the best data you can get (AIS and radar to verify actual range), and head off at the last moment to avoid getting under his bows, which will be hellishly difficult to judge and really dangerous. Then turn hard to run down his side, and try to judge when you can turn to pass behind him, without banging into his quarter -- really hard when your combined speeds are 25 knots (!).

This is totally unseamanlike, violating several COLREGS, but it's the ONLY way you can get within a cable of his quarter with a 4:1 speed differential.

The seamanlike way is to set up a CPA from a couple of miles out which is large enough to be able to measure -- large enough to be a meaningful, correctable number. Get onto that glide path early and hold your course while closely monitoring AIS and radar and making corrections when it's possible -- when you have meaningful data.

I say again, when he's running at you at 4 times your speed, less than two points off dead abeam, from a mile or two away, your eyeball tells you NOTHING. A couple of degrees plus or minus will put you under his bows, on the one hand, or give you a missed approach, on the other, where you end up too far away to get back to where you could pass safely in front of the next ship, and you cannot see the difference between 16 degrees and 12 degrees from dead abeam of you, with your bare eyes, nor do you even know within a point or even two, how close to perpendicular his course is to you. You cannot do this at all if you don't understand the limitations of what you can see and perceive.

What you propose to do is like trying to land plane by flying really close to the ground, and then slamming the stick forward at the moment you glimpse the runway. Thinking that by approaching from much closer to the ground, you can land in a shorter distance. It doesn't work like that. If you are trying to land on a really short runway, you need all the more to approach from the correct glide path, in order to see the spot you are going to land from further away, and have more control over getting there. You don't aim for a spot which is three feet from the fence -- you aim for a spot far enough from the fence that it is within what you can reasonably know, so that your corrections are small and not desperate and last minute.

Just the same in this situation -- if you set up from a couple miles away a crossing with a CPA which is within what you can meaningfully measure, then you actually have a chance to make corrections, and you actually have a chance of passing somewhere near the point you aimed at. Instead of wild maneuvers, done by eye, at the last second, which are just as likely to put you under the bows of the next ship as anything.


The biggest ship in the world, travelling at 20 knots, will be GONE and past you, in less than 30 seconds from the time his bow goes past you, to the time his transom goes past you -- 1/5 of a nautical mile of ship flying past as you bob around less than a cable away. You can only move maximum 300 feet in that amount of time, at 5 knots. See Lodesman's plot of this -- in order to pass 180 feet from his quarter, you have to be already LESS THAN A CABLE from his bow, then as his bow flashes past, you have to keep running at top speed directly at him, during the 30 seconds it takes for him to flash past. How little you visualize this is evident when you talk about "doubling back" -- you can't get there, from there. He's not anchored -- and when you get close, it looks like a bullet going by.
Buddy, all I can tell you, is that you need to get on a race course and practice crossing vessels in close quarters to get any clue what you are talking about.

Keeping you eyes on the radar or AIS or whatever will get you smacked.

Once you get inside the useful range of your electronics, it all has to be done by eyeball.

Anyone attempting this kind of crossing in anything less than clear visibility has a death wish.

Therefore, I completely stand behind my post # 591.

Any attempt at this crossing has risk (as does any kind of sailing). Some may even say attempting this crossing is unsafe under any circumstance. Others who have a lot of experience crossing ships in close quarters may be sufficiently comfortable (I don't mean relaxing with a beer, but are confident in their ability to avoid collision).

Regardless of the individuals Pucker Factor, and anyones opinion whether attempting this crossing is safe on any terms, if one proceeds, crossing with a CPA of 180 ft from the lead ship port transom corner, is more safe, than crossing at 1/4 nm CPA.

Period.

Nothing you can say, will change that, because it is absolutely, 100% correct, as proven with my post # 591 calcs.
ramblinrod is offline   Reply With Quote
Old 24-10-2017, 11:36   #597
cruiser

Join Date: Jan 2012
Location: Lake Ontario
Boat: Ontario 38 / Douglas 32 Mk II
Posts: 3,250
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by Dockhead View Post

The key here is the phrase "maintain" a 1/4 mile CPA. You can't maintain it.
Focusing a little more attention on this completely incorrect statement, if one can't maintain a CPA value, then WTF is the point of even considering a solution, or even attempting to navigate for that matter?

The answer is, OF COURSE one can maintain a CPA.

That is what one does in every crossing. Determine the minimum CPA from the other vessel, and then makes dang sure they maintain it.

If you can't maintain a minimum CPA, you very well may end up under him as shark bait. So long ol' chum.
ramblinrod is offline   Reply With Quote
Old 24-10-2017, 11:40   #598
cruiser

Join Date: Jan 2012
Location: Lake Ontario
Boat: Ontario 38 / Douglas 32 Mk II
Posts: 3,250
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
Sensible.

I liken the safety zone around a vessel like a tear drop.

The closer to the transom, the nearer the "safe distance".

The further from the transom, the greater the "safe distance".

There is a minimum distance, as one needs to contend with variations in wind, waves, current, visibility, etc. The greater the possible error induced by the variables, the greater the safe distance.

I completely concur, for the given scenario, the further distance one can maintain from the following ship, the better, and if that means getting very close to the port transom corner of the lead ship, sobeit.

Will there be dirty shorts on the lead ship? Possibly.

Will the sailboat skipper be on high alert, maintaining a watchful eye on the lead ship during the entire crossing? Sure better.

Should he be dicking with AIS, Radar, GPS, and Marpa AFTER he is in close quarters with the ship and controlling the crossing by eyeball? Not on his life!
Correction: typo, fourth sentence should have read...

The further (ahead of) the transom, the greater the "safe distance".
ramblinrod is offline   Reply With Quote
Old 24-10-2017, 11:59   #599
Moderator
 
Paul Elliott's Avatar

Cruisers Forum Supporter

Join Date: Sep 2006
Posts: 4,601
Images: 4
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
Buddy, all I can tell you, is that you need to get on a race course and practice crossing vessels in close quarters to get any clue what you are talking about.
Rod, this is not a useful comparison. I occasionally race, and according to the RRS (Racing Rules of Sailing) sometimes a miss *is* *not* as good as a mile. If you fail to give room at the mark and so impede a sailboat from it's "proper course" you have fouled that boat. A collision is not required.

When racing all the participants have agreed to abide by the RRS. We are all sailing boats of a similar size (+/- 50 ft or so), at similar speeds (+/- 5 kts), and close crossings are expected. Colllisions are seldom life-threatening.

That 600-ft ship has not agreed to follow the RRS. It is sailing according to COLREGS. It can't maneuver like a racing sailboat, and it is usually traveling *much* faster than you are. It isn't out there to have fun, but to safely carry passengers and cargo.
__________________
Paul Elliott, S/V VALIS - Pacific Seacraft 44 #16 - Friday Harbor, WA
www.sailvalis.com
Paul Elliott is offline   Reply With Quote
Old 24-10-2017, 12:21   #600
Moderator
 
Dockhead's Avatar

Cruisers Forum Supporter

Join Date: Mar 2009
Location: Denmark (Winter), Helsinki (Summer); Cruising the Baltic Sea this year!
Boat: Cutter-Rigged Moody 54
Posts: 28,478
Re: Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

Quote:
Originally Posted by ramblinrod View Post
Buddy, all I can tell you, is that you need to get on a race course and practice crossing vessels in close quarters to get any clue what you are talking about.
I have explained in great detail why it is completely different, when the speed and size of the vessels are different, and especially when you have a large difference of speed like this.

You are applying a completely different set of parameters which cannot be applied to crossing with fast ships at sea. You can eyeball a crossing pretty well when two slow vessels are approaching each other at a decent angle. You cannot do it when you're dealing with a high speed, large vessel, approaching you from nearly your beam.

You are imagining mixing up with fast ships at sea like you jostle with race boats around the cans -- it doesn't work like that at all! It's a pity we don't have a simulator -- you would see it immediately. It can be shown with math -- starting with angles which are visible and invisible to the human eye at what distance, from which we can get what rate of bearing change is visible and what is not perceptible, then how accurately distances can be judged. From that we can show that you can't, for example, tell the difference between passing three cables ahead from passing three cables behind, from a mile out, at these speeds -- something I know from experience, but can also be demonstrated by math. This would be a really useful exercise, actually. I am working this week and don't have time for this, but maybe our math whizz Juho would like to have a crack at it.
__________________
"Parce que je suis heureux en mer, et peut-ętre pour sauver mon ame. . . "
Dockhead is online now   Reply With Quote
Reply

Tags
collision

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Challenge: Collision Avoidance! Pelagic Challenges 53 18-08-2017 19:54
CARD Collision Avoidance Radar Detector multihullsailor6 Marine Electronics 12 27-12-2015 21:12
Collision Avoidance - Tsunami Debris rreeves Health, Safety & Related Gear 22 03-05-2012 07:23
Collision Avoidance in Mexico: AIS or Radar or ? no_bad_days Pacific & South China Sea 27 19-09-2011 15:40
Distance to Horizon & Collision Avoidance GordMay General Sailing Forum 7 19-06-2009 00:18

Advertise Here


All times are GMT -7. The time now is 09:21.


Google+
Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Social Knowledge Networks
Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2020, vBulletin Solutions, Inc.

ShowCase vBulletin Plugins by Drive Thru Online, Inc.