Excellent thing for physics simulation in space. Just, in UE5.1 if you add cube, then it will not allow you to set 'Simulate Physics' because of missing collision structure. To rectify that, in Outliner select cube, right click and 'Edit Cube', then under 'Collision' click on 'Add Box Simplified Collision'. Now it is possible to enable 'Simulate Physics'. Another problem remains with accelerating of the objects. Tried to increase 'Linear Damping' as well as 'Angular Damping', and got somewhat stable configurations unless some of the objects has too big mass. Then other objects with smaller mass being propelled to the space.
Hello, i don't know if you still answer but i'll try my luck^^ i have two questions, do you have a way to make a kind of solar system like you show by the end (i know that decreasing the G constant and making bigger masses will arrange that but do you have any tip ? i already understood that you used the freeze on XYZ of the sun but for the rest i'm lost and don't know how to calculate that) And how did you set up the trail behind the planet (like you show at the end) ? I would be really grateful if you could help me (as if i wasn't already with this tutorial ^^)
This is really fascinating, thank you. I might try to have a go at it myself. Can you explain when, after reaching a certain spin velocity, the objects start getting ejected? like @11:39
Even though objects are already stuck to each other, they still apply forces on each other. Sometimes that results in a force that spins the entire thing. Since we re in vacuum, it keeps on accelerating, perhaps the movements of magma inside the earth are cause by a similar force. The more it spins the more the objects are pushed outwards, it’s the centrifugal force, have a google. At some point the gravity pushing the objects to the surface becomes smaller than the force pushing outwards and the object takes off. After that all depends on the speed. If the objects is fast enough it will escape the gravity of the rest completely and leave, if not it will start orbiting and fall back down again at some point.
@@devlogsadgi6778 Thanks for the reply! I'm aware it might be centrifugal force, I'm just amazed that it was also calculated by the physics system of Unreal.:D I'm still a newb in Unreal, and dont know much about the physics system.. I thought it would calculate and apply only what was initiated in the Main blueprint (which was all attractive force). Thanks again!
This is a really awesome tutorial! As a noobie I'm really struggling to follow because using shortcuts just brings up windows like magic. I'm stuck on how you brought up graph editor because the tabs just changed when you imported the sky box....
Just to add the event graph has changed considerably now in unreal 5.1, with length array not existing / moved to something else and actor arrays now givng an extra "make array" element that I'm not sure makes a difference. Edit: my bad, it was because I tried creating an element by dragging from the actors output (it doesn't include length for some reason)
@@Captain-Comedy I had the same issue. The reason is that we haven't changed the masses and actors into array type when we created them. Look at the top right, next to variable type there is another option to make it array.
This video is amazing. But when I followed the video, I found some questions. Initial velocity (calculated) is far away not enough in simulation. there seemed a constant changing with planet itself mass and constant G and center object mass and orbital height. The second thing is seemed orbital can never stay stable. There is always a moment that planet will fly away or drop in. It's there any way to fix? And that "freeze" problem is ue5 bug. when speed is low or added force is small, the editor will put actor into sleep in error. Hope they can fix it in ue5.1
There is a way to keep them in orbit better, it’s to go more realistic on the masses and G constant. Lower G way down, and increase mass way up. The orbits will stabilize
I’m wondering how this compares to my effort created in Unity. I wanted at least hundreds of objects. Using the Unity job system, burst compiler, it will smoothly handle up to about 4,000 low poly objects (spheres take to long to render). It’s very code oriented, and I’ve included options for start conditions, and G can be changed as the simulation is running. I’m intending to recreate it in Unreal as a learning project. Doubt this method will work for thousands of objects but interesting nonetheless.
How do I prevent the game from trowing that error "Attempted to access StaticMeshActor_UAID_00D861553EFF651101_2006019908 via property CallFunc_Array_Get_Item_3, but StaticMeshActor_UAID_00D861553EFF651101_2006019908 is not valid (pending kill or garbage)" It seems like its the spheres that get lost into that glitch that cause this how do I prevent this I keep trying with an is valid node but its not preventing the error it tanks my fps. it says its the Add force fault and casting to static mesh taht doesnt exist i guess. I dont get how to fix I tried deleting them and checking validity
I hope you will respond even if it's been a long time since this video was posted. I'm not yet an expert on blueprints and I wanted to ask you how I can modify the nodes, to exclude objects with negligible mass among them. Something like: if object A and object B have a mass < 10, then don't apply the forces between them, but have them attracted by object C which has a mass > 10. I hope my question is clear. Thank you!
Hey, so check where i am creating the event tick function yes? I access the mass array twice using get there, connect both to an add function to sum those to values, then use a bigger than node to check if the sum is bigger then 10, connect that boolean result to a branch node, that branch node should go right before the casttostaticmeshcomponent node, now if TRUE output should connect to casttostaticmeshcomponent and the following add force node, and you can leave the FALSE output empty. I haven’t tested, but seems to me that should skip applying forces to objects that have cumulative mass of less or equal than 10. Hopefully that should be enough to get you going, otherwise ask me again.
However if the masses change while simulation is running you might start experiencing objects that stop moving, or start going. So watch out, but if masses are stable then you should have only big masses affecting the system, perhaps making orbits for stable and predictable. If that is your goal, then i advise to greatly reducing constant G value and greatly increasing masses, it will stabilise orbits.
I am getting an error when I try this "Blueprint Runtime Error: "Attempted to access index 2 from array Masses of length 2!" Node: Add Force Graph: ..." Any idea what I am doing wrong?
hey, the celestial bodies just stop moving if they're too far away from themselfes, is there a way to fix it? And also how do you find out the velocity you need for the planets to orbit themselfes?
I mention the issue of objects stopping to move when far away at the end of the video. This bug only occurs in UE5, not in UE4, my guess is that the UE5 release isn’t stable enough yet. You could use UE4, same script works there. And for the orbital velocity, well you could eye ball it or use a formula such as this: getcalc.com/physics-orbital-velocity-calculator.htm
@@devlogsadgi6778 a bit more is that sometimes Chais put objets in sleep mode to not calculate physic everywhere, there is some sleep parameters we can set up but I did not try maybe it will be worth the attempt :) there is a sleep family and also a float about sleep
I was able to figure out a fix for the sleeping physics at larger distances. If you assign a physics material and set the linear and angular sleep thresholds to 0 it fixes this. Alternatively, using the command p.Chaos.Solver.SleepEnable 0 fixes this as well, although this might be editor only and not persist to an actual build
@@cl0useau949 Thanks! The command works for the editor but if instead of "Selected Viewport" you click play on "Standalone Game" it no longer works which confirms your hypothesis about it being editor only. Noob question though, how do I assign a physics material and set those thresholds? I can't seem to find any physics material or such thresholds.
Well, it sounds like you just need to fine tune the force of gravity. You could do so by decreasing the G constant or the planets masses. Also, for a stable system the centre body’s mass should exponentially exceed others. Otherwise it will move too freely and collide with other objects, ceasing the orbits. I personally chose to simply constraint the sun’s x/y/z displacements.
@@devlogsadgi6778 Thank you i will try what you said. one more thing, do you think there is there anyway to show a orbit path before you run the game or no? thank you for your help.
@@bigsk8743 I have considered this, however the system as it is proves itself being inconsistent. The same values of masses/velocities/everything else yields different orbits on different plays. Which is likely caused by inconsistent rounding by the compiler. Given that fact predicting an orbit for a long time is not feasible. So I would say, the only dish on the menu in terms of predicting is the next 5-10 seconds of movement, even this would yield some significant level of inaccuracy. In terms of implementing that, it wouldn’t be hard, you just have to create a script that iterates a certain amount of times the current values in the system, but instead of applying them to object just saves them for the next iteration. Then you will end up with an array positions for n interactions. All you do next is create some particles or just a small sphere at those coordinates. You will end up with a angry birds type of prediction. If you set it up in the constructor than it will display that prediction even in the editor. Let me know if you have any other questions, glad to brainstorm.
@@devlogsadgi6778 thank you I'm working with slow orbits and I was trying to avoid wasting time tweaking over and over, thank you so much for your help :)
Hey! I created an Actor Blueprint, I was setting up everything (settings, root element, tag as well), but the spheres won't move at all... What did I do wrong? :/
It is surely possible, but it would take time to give the user the ability to create their own setup with their settings. I am not going to go that way just yet))
Concise, extremely useful and wonderful-looking graphs! Great how realistic the behaviour in the ending montage is!
Very concise Tutorial. Cheers, matey!
Sigue así bro, estos videos molan
Perfect tutorial! I've made orbit simulations in 2D before, but this was the perfect intro for simulating in 3D, thanks heaps
Exellent video!
I understanded half of it but i subscribed anyway because it was so much interested in the video i could not resist.
I’ve been trying to do something similar with c++ and I just couldn’t get it work for some reason, now I finally got a solution thanks
Excellent thing for physics simulation in space. Just, in UE5.1 if you add cube, then it will not allow you to set 'Simulate Physics' because of missing collision structure. To rectify that, in Outliner select cube, right click and 'Edit Cube', then under 'Collision' click on 'Add Box Simplified Collision'. Now it is possible to enable 'Simulate Physics'. Another problem remains with accelerating of the objects. Tried to increase 'Linear Damping' as well as 'Angular Damping', and got somewhat stable configurations unless some of the objects has too big mass. Then other objects with smaller mass being propelled to the space.
Excelente video
Thank you so much
good! 구독했습니다
How did you make the speed impulse to generate orbits? Sorry, this is my first experience with UE5
Use the Add Force node to the object, I briefly mention it at 14:46
Hello, i don't know if you still answer but i'll try my luck^^
i have two questions, do you have a way to make a kind of solar system like you show by the end (i know that decreasing the G constant and making bigger masses will arrange that but do you have any tip ? i already understood that you used the freeze on XYZ of the sun but for the rest i'm lost and don't know how to calculate that)
And how did you set up the trail behind the planet (like you show at the end) ?
I would be really grateful if you could help me (as if i wasn't already with this tutorial ^^)
This is really fascinating, thank you. I might try to have a go at it myself. Can you explain when, after reaching a certain spin velocity, the objects start getting ejected? like @11:39
Even though objects are already stuck to each other, they still apply forces on each other. Sometimes that results in a force that spins the entire thing. Since we re in vacuum, it keeps on accelerating, perhaps the movements of magma inside the earth are cause by a similar force.
The more it spins the more the objects are pushed outwards, it’s the centrifugal force, have a google. At some point the gravity pushing the objects to the surface becomes smaller than the force pushing outwards and the object takes off. After that all depends on the speed. If the objects is fast enough it will escape the gravity of the rest completely and leave, if not it will start orbiting and fall back down again at some point.
@@devlogsadgi6778 Thanks for the reply! I'm aware it might be centrifugal force, I'm just amazed that it was also calculated by the physics system of Unreal.:D I'm still a newb in Unreal, and dont know much about the physics system.. I thought it would calculate and apply only what was initiated in the Main blueprint (which was all attractive force). Thanks again!
This is a really awesome tutorial!
As a noobie I'm really struggling to follow because using shortcuts just brings up windows like magic. I'm stuck on how you brought up graph editor because the tabs just changed when you imported the sky box....
Just to add the event graph has changed considerably now in unreal 5.1, with length array not existing / moved to something else and actor arrays now givng an extra "make array" element that I'm not sure makes a difference.
Edit: my bad, it was because I tried creating an element by dragging from the actors output (it doesn't include length for some reason)
@@Captain-Comedy I had the same issue. The reason is that we haven't changed the masses and actors into array type when we created them. Look at the top right, next to variable type there is another option to make it array.
I don't understand how that freezing sphere glitch is still a glitch?
This video is amazing. But when I followed the video, I found some questions. Initial velocity (calculated) is far away not enough in simulation. there seemed a constant changing with planet itself mass and constant G and center object mass and orbital height. The second thing is seemed orbital can never stay stable. There is always a moment that planet will fly away or drop in. It's there any way to fix?
And that "freeze" problem is ue5 bug. when speed is low or added force is small, the editor will put actor into sleep in error. Hope they can fix it in ue5.1
There is a way to keep them in orbit better, it’s to go more realistic on the masses and G constant. Lower G way down, and increase mass way up. The orbits will stabilize
I’m wondering how this compares to my effort created in Unity. I wanted at least hundreds of objects. Using the Unity job system, burst compiler, it will smoothly handle up to about 4,000 low poly objects (spheres take to long to render). It’s very code oriented, and I’ve included options for start conditions, and G can be changed as the simulation is running.
I’m intending to recreate it in Unreal as a learning project. Doubt this method will work for thousands of objects but interesting nonetheless.
How do I prevent the game from trowing that error "Attempted to access StaticMeshActor_UAID_00D861553EFF651101_2006019908 via property CallFunc_Array_Get_Item_3, but StaticMeshActor_UAID_00D861553EFF651101_2006019908 is not valid (pending kill or garbage)" It seems like its the spheres that get lost into that glitch that cause this how do I prevent this I keep trying with an is valid node but its not preventing the error it tanks my fps. it says its the Add force fault and casting to static mesh taht doesnt exist i guess. I dont get how to fix I tried deleting them and checking validity
Is valid node
I hope you will respond even if it's been a long time since this video was posted. I'm not yet an expert on blueprints and I wanted to ask you how I can modify the nodes, to exclude objects with negligible mass among them. Something like: if object A and object B have a mass < 10, then don't apply the forces between them, but have them attracted by object C which has a mass > 10. I hope my question is clear. Thank you!
Hey, so check where i am creating the event tick function yes? I access the mass array twice using get there, connect both to an add function to sum those to values, then use a bigger than node to check if the sum is bigger then 10, connect that boolean result to a branch node, that branch node should go right before the casttostaticmeshcomponent node, now if TRUE output should connect to casttostaticmeshcomponent and the following add force node, and you can leave the FALSE output empty. I haven’t tested, but seems to me that should skip applying forces to objects that have cumulative mass of less or equal than 10. Hopefully that should be enough to get you going, otherwise ask me again.
However if the masses change while simulation is running you might start experiencing objects that stop moving, or start going. So watch out, but if masses are stable then you should have only big masses affecting the system, perhaps making orbits for stable and predictable. If that is your goal, then i advise to greatly reducing constant G value and greatly increasing masses, it will stabilise orbits.
Is it possible to cast to a blueprint too? and if so how?
I am getting an error when I try this "Blueprint Runtime Error: "Attempted to access index 2 from array Masses of length 2!" Node: Add Force Graph: ..." Any idea what I am doing wrong?
hey, the celestial bodies just stop moving if they're too far away from themselfes, is there a way to fix it? And also how do you find out the velocity you need for the planets to orbit themselfes?
I mention the issue of objects stopping to move when far away at the end of the video. This bug only occurs in UE5, not in UE4, my guess is that the UE5 release isn’t stable enough yet. You could use UE4, same script works there.
And for the orbital velocity, well you could eye ball it or use a formula such as this:
getcalc.com/physics-orbital-velocity-calculator.htm
@@devlogsadgi6778 a bit more is that sometimes Chais put objets in sleep mode to not calculate physic everywhere, there is some sleep parameters we can set up but I did not try maybe it will be worth the attempt :) there is a sleep family and also a float about sleep
I was able to figure out a fix for the sleeping physics at larger distances. If you assign a physics material and set the linear and angular sleep thresholds to 0 it fixes this. Alternatively, using the command p.Chaos.Solver.SleepEnable 0 fixes this as well, although this might be editor only and not persist to an actual build
@@cl0useau949 Thanks! The command works for the editor but if instead of "Selected Viewport" you click play on "Standalone Game" it no longer works which confirms your hypothesis about it being editor only. Noob question though, how do I assign a physics material and set those thresholds? I can't seem to find any physics material or such thresholds.
hey ive been trying to set up orbits, and i give the planets an initial velocity, but the orbits decay and crash into each other so quick any tips?
Well, it sounds like you just need to fine tune the force of gravity. You could do so by decreasing the G constant or the planets masses. Also, for a stable system the centre body’s mass should exponentially exceed others. Otherwise it will move too freely and collide with other objects, ceasing the orbits. I personally chose to simply constraint the sun’s x/y/z displacements.
@@devlogsadgi6778 Thank you i will try what you said. one more thing, do you think there is there anyway to show a orbit path before you run the game or no? thank you for your help.
@@bigsk8743 I have considered this, however the system as it is proves itself being inconsistent. The same values of masses/velocities/everything else yields different orbits on different plays. Which is likely caused by inconsistent rounding by the compiler. Given that fact predicting an orbit for a long time is not feasible. So I would say, the only dish on the menu in terms of predicting is the next 5-10 seconds of movement, even this would yield some significant level of inaccuracy. In terms of implementing that, it wouldn’t be hard, you just have to create a script that iterates a certain amount of times the current values in the system, but instead of applying them to object just saves them for the next iteration. Then you will end up with an array positions for n interactions. All you do next is create some particles or just a small sphere at those coordinates. You will end up with a angry birds type of prediction. If you set it up in the constructor than it will display that prediction even in the editor. Let me know if you have any other questions, glad to brainstorm.
@@devlogsadgi6778 thank you I'm working with slow orbits and I was trying to avoid wasting time tweaking over and over, thank you so much for your help :)
@@devlogsadgi6778 hi there, sorry for silly question, how do I change the constant g for individual planets?
wow
Glad you liked it!
Hey!
I created an Actor Blueprint, I was setting up everything (settings, root element, tag as well), but the spheres won't move at all... What did I do wrong? :/
Same!! I have this problem too :( Did you figure this out?
@@SpandaV2 Not yet.
gj
can you export this like a game to play
It is surely possible, but it would take time to give the user the ability to create their own setup with their settings. I am not going to go that way just yet))