I am trying to construct a Convergent Tetra Spade interchange; still . . .

[Pythias900KMB]

Many salutations, everybody!  As you can see from the title and from the attached screen shots, I am hard at work revising an earlier RHW interchange galvanized by my newly acquired insight about how to avail the FlexFly; unfortunately, it is not going well.  The western approach to the interchange has a railway crossing it; while I have been able to thread the railway to run under the RHW; the game will throw a major temper tantrum if I try to do that with GHSR.  I had to do a bit of gymnastics to revise the OST so that I could add in a pair of inside A1 ramps.  I have also had to go old school with the right-turn ramps  . . . well, you can see for yourselves how that has gone.  One of the left-turn ramps keeps overriding back to RHW-2 against my wishes . . . guys, help!

[Tarkus]

The adjacency stability code to do what you want doesn't exist, and it's a little hard to tell, but the first might entail a three-level crossing.  The tighter the interchange is, the more adjacency stability is needed, and it gets really hard to predict exactly what everyone will do.

-Alex

[Pythias900KMB]

The adjacency stability code to do what you want doesn't exist, and it's a little hard to tell, but the first might entail a three-level crossing.  The tighter the interchange is, the more adjacency stability is needed, and it gets really hard to predict exactly what everyone will do.

-Alex

Gratitude for the insight about the adjacency stability code, Tarkus; still, that is not going to get this interchange finished.  How do I use that data when I could not even avail the FlexFly for the right-turn ramps at all 

[Tarkus]

The fact that those ramps are reverting to RHW-2 suggests you're not actually using FLEXFly, but the Multi-Radius Curves (MRCs) instead.  Indeed, those RHW-4 ramps you have there are actually less smooth than the FLEXFly pieces, to boot.  The actual FLEXFly pieces are either fully MIS or RHW-4-based, and the curve network is not overrideable--the overrideable portion is the under/overcrossing functionality.  By comparison, the MRCs are RHW-2-based, and the curve network can be overridden, but they can't accept under/overcrossings. 

Back to the adjacency stability situation, if the code doesn't exist for the particular situations you're trying to build, unfortunately, there's not really a way around that, other than making the interchange less compact, such that there's at least one tile of base orthogonal or diagonal.  Those base adjacencies tend to be relatively universal, especially in comparison to some of the ones that have come about in your current attempt (though with the codebase being what it is, I can't guarantee that there might not be some weird issue with things getting flipped around the opposite way of what you want in certain situations).

-Alex

[Pythias900KMB]

The fact that those ramps are reverting to RHW-2 suggests you're not actually using FLEXFly, but the Multi-Radius Curves (MRCs) instead.  Indeed, those RHW-4 ramps you have there are actually less smooth than the FLEXFly pieces, to boot.  The actual FLEXFly pieces are either fully MIS or RHW-4-based, and the curve network is not overrideable--the overrideable portion is the under/overcrossing functionality.  By comparison, the MRCs are RHW-2-based, and the curve network can be overridden, but they can't accept under/overcrossings. 

Back to the adjacency stability situation, if the code doesn't exist for the particular situations you're trying to build, unfortunately, there's not really a way around that, other than making the interchange less compact, such that there's at least one tile of base orthogonal or diagonal.  Those base adjacencies tend to be relatively universal, especially in comparison to some of the ones that have come about in your current attempt (though with the codebase being what it is, I can't guarantee that there might not be some weird issue with things getting flipped around the opposite way of what you want in certain situations).

-Alex

Alex,

The left-turn ramp carrying traffic toward the south and the left-turn ramp carrying traffic toward the west have a 1st elevation lvl viaduct involved -- the latter being because of a rail line.  Incidentally, is there a way to lower from 2nd lvl to 1st lvl elevation a draggable rail viaduct in the same manner as the RHW?  If that capability existed, I would not have had such a headache with the western approach.

I believe what would solve this problem is if each egress ramp could diverge in both directions like is the case with RL highway interchanges.  I tried that with the western approach only for the D2 ramp to remain at ground lvl even though I availed an L2 RHW-4S starter piece.  Another problem lies with merging the ramps back into the main highway; what would need to happen if I avail an A2 egress ramp is something like this scenario:  If you are heading west on the McDermott Freeway (IH-10) into Balcones Heights and take egress #564 to turn west onto the Connally Loop (Loop 410), you will find that the two lanes of the ramp will be added alongside the main lanes thus resulting in a six lane carriageway -- the outer lane of which is forced to egress at Babcock Road.  I really would like explained why there are motorists that get confused -- and cause a serious traffic hazard -- over whether to avail or bypass an interchange.  It is not like there are no signs giving them prior notice of the interchange or directing them to which lanes avail the interchange along with telling them where they will end up if they diverge to the right or to the left.  Instead of all these fancy highway interchanges, I would rather have motorists of much greater competency and attentiveness . . .

Anyway, back to this.  So your advice is to space out the ramps a bit so that the RUL files have an easier time sorting themselves out?  I already avail a two-tile median as standard operating procedure, so if I want to cross two left-turn ramps, the actual junction has to be in that median.  Due to the rail line, I have had to construct the ramps on viaducts -- though if I could lower the railway to L1 and have that left-turn ramp to the west cross under it while the east-west highway crosses over it, that would help tremendously.  So is that possible?

This is why I was asking about the RHW-12S, Alex;

In contrast, that is not possible with the Network Amendment Module the way it currently is