Difference between revisions of "Jo Blocks"
(→Motivation: +) |
(→Method: +) |
||
| (One intermediate revision by the same user not shown) | |||
| Line 28: | Line 28: | ||
The system employed by KSP in VAB to assemble complete craft from parts requires that a ''tree'' be formed: '''Each part must attach to one and only one parent part.''' From the viewpoint of mechanical engineering this introduces a serious weakness in many models, since the conventional method of building large, mechanically stable structures is to join each part to several neighbors, forming loops. Indeed it is difficult to find any real, extensive structure in which there are not multiple connections. However, in KSP, loops in the part network are forbidden. | The system employed by KSP in VAB to assemble complete craft from parts requires that a ''tree'' be formed: '''Each part must attach to one and only one parent part.''' From the viewpoint of mechanical engineering this introduces a serious weakness in many models, since the conventional method of building large, mechanically stable structures is to join each part to several neighbors, forming loops. Indeed it is difficult to find any real, extensive structure in which there are not multiple connections. However, in KSP, loops in the part network are forbidden. | ||
| − | The stock game provides one obvious workaround to this limitation: the [[EAS-4 Strut Connector]] or '''strut'''. Struts are peculiar in that the far end point and the span between ends are "not really there"; the far end is recalculated when a craft is edited. When subassemblies are cloned, stored, or recalled; when the craft is re-rooted with [[SelectRoot]]; when symmetry is employed or changed; | + | The stock game provides one obvious workaround to this limitation: the [[EAS-4 Strut Connector]] or '''strut'''. Struts are peculiar in that the far end point and the span between ends are "not really there"; the far end is recalculated when a craft is edited. When subassemblies are cloned, stored, or recalled; when the craft is re-rooted with [[SelectRoot]]; when symmetry is employed or changed; or when parts of different sizes are substituted; there is a risk that existing struts will fail to attach as intended. Struts may be somewhat difficult to attach or seem to be required in large numbers. Also, the visual appearance of struts may be offensive in some places. Despite these drawbacks, struts are often the best choice and few large craft are entirely without them. |
| − | An alternative method of forming an effective part network loop is to employ | + | An alternative method of forming an effective part network loop is to employ [[Clamp-O-Tron Docking Port]]s as '''paired ports'''. If these are assembled carefully then one port of each pair attaches normally to the part tree; they do ''not'' attach to one another until the craft is constructed (launched on pad). At this time the pair immediately dock, as they are very close and correctly aligned; and close the mechanical loop. |
| − | A further consideration is flexibility. Strut connections are inflexible, hence brittle; they will absorb considerable load but then fail. Docked ports are more flexible and will yield under stress | + | A further consideration is flexibility. Strut connections are inflexible, hence brittle; they will absorb considerable load but then fail. Docked ports are more flexible and will yield under stress. If the ports are not pulled completely into an undocked state then, when the stress is relieved, they will relax into their former neutral state. |
Key to the paired-port strategy is correct assembly. If parts overlap, even slightly, then "ghost forces" arise that cause uncontrolled attitude changes, even changes in velocity vector. If the overlap is greater then the parts involved, even the entire craft, may be totally destroyed at once. ''(See [[Kraken]].)'' If the gap between paired ports is too large then, as they pull together in an attempt to dock, they will apply stresses to the rest of the craft; if they are out of alignment then the stresses may be more serious. '''The ideal pairing leaves a scant millimeter of free space between ports.''' If you cannot see between the ports then they may well be overlapping. | Key to the paired-port strategy is correct assembly. If parts overlap, even slightly, then "ghost forces" arise that cause uncontrolled attitude changes, even changes in velocity vector. If the overlap is greater then the parts involved, even the entire craft, may be totally destroyed at once. ''(See [[Kraken]].)'' If the gap between paired ports is too large then, as they pull together in an attempt to dock, they will apply stresses to the rest of the craft; if they are out of alignment then the stresses may be more serious. '''The ideal pairing leaves a scant millimeter of free space between ports.''' If you cannot see between the ports then they may well be overlapping. | ||
| Line 39: | Line 39: | ||
== Method == | == Method == | ||
| + | |||
| + | Talisar's structural truss is available in [http://forum.kerbalspaceprogram.com/threads/30673-0-23-5-Spherical-and-Toroidal-Tank-Pack-%28Updated-05-02-14%29-%28New-download-link%29 Spherical and Toroidal Tank Pack] (recommended in any case) and will be used throughout as a basis of measurement. Trial craft are built in VAB with no intention of flight, merely to compare lengths. The shortest truss is 0.5 m, although there is a mini truss with 0.25 m parts. The standard 6-way hub is also one meter between opposing faces. All truss parts stack properly: a stack of ten hubs is the same length as one 10 m truss part. | ||
| + | |||
| + | What to do when a part does not measure an exact multiple of known parts? The method of solution is to multiply instances of the part to be measured until the whole stack reaches a measurable length. For example, one Toroidal Tank Hub (2.5m), also from Talisar, is less than 0.5 m and so cannot be measured. But fifty such parts can be measured to 10 m; hence each part is 0.20 m (long or thick). Future measurements may now employ the newly measured part. | ||
Latest revision as of 18:21, 7 July 2014
Jo blocks (Johansson gauges, gauge blocks) are, in KSP, a concept rather than a mod or set of physical parts. Real Jo blocks are small rectangular precision-machined parts that are stacked together or wrung to make up arbitrary lengths. Ingame, every instance of each part is precisely as long as every other instance; but it is not always immediately clear what that length is.
Definitions
- Length
- Distance between stack nodes, irrespective of model rendering. Inapplicable to surface attachment.
- One meter
- Defined here as the length of one instance of Taliar's (Standard) Cubic Truss (1m) as of 2014 Jly 01.
- Wringing
- Stacking a number of parts without possibility of surface attachment using Editor Extensions to disable.
Motivation
The system employed by KSP in VAB to assemble complete craft from parts requires that a tree be formed: Each part must attach to one and only one parent part. From the viewpoint of mechanical engineering this introduces a serious weakness in many models, since the conventional method of building large, mechanically stable structures is to join each part to several neighbors, forming loops. Indeed it is difficult to find any real, extensive structure in which there are not multiple connections. However, in KSP, loops in the part network are forbidden.
The stock game provides one obvious workaround to this limitation: the EAS-4 Strut Connector or strut. Struts are peculiar in that the far end point and the span between ends are "not really there"; the far end is recalculated when a craft is edited. When subassemblies are cloned, stored, or recalled; when the craft is re-rooted with SelectRoot; when symmetry is employed or changed; or when parts of different sizes are substituted; there is a risk that existing struts will fail to attach as intended. Struts may be somewhat difficult to attach or seem to be required in large numbers. Also, the visual appearance of struts may be offensive in some places. Despite these drawbacks, struts are often the best choice and few large craft are entirely without them.
An alternative method of forming an effective part network loop is to employ Clamp-O-Tron Docking Ports as paired ports. If these are assembled carefully then one port of each pair attaches normally to the part tree; they do not attach to one another until the craft is constructed (launched on pad). At this time the pair immediately dock, as they are very close and correctly aligned; and close the mechanical loop.
A further consideration is flexibility. Strut connections are inflexible, hence brittle; they will absorb considerable load but then fail. Docked ports are more flexible and will yield under stress. If the ports are not pulled completely into an undocked state then, when the stress is relieved, they will relax into their former neutral state.
Key to the paired-port strategy is correct assembly. If parts overlap, even slightly, then "ghost forces" arise that cause uncontrolled attitude changes, even changes in velocity vector. If the overlap is greater then the parts involved, even the entire craft, may be totally destroyed at once. (See Kraken.) If the gap between paired ports is too large then, as they pull together in an attempt to dock, they will apply stresses to the rest of the craft; if they are out of alignment then the stresses may be more serious. The ideal pairing leaves a scant millimeter of free space between ports. If you cannot see between the ports then they may well be overlapping.
Clearly, for those wishing to pursue paired ports, it's required to know precisely the length of all parts involved in a loop. It should go without saying that surface attachment must be excluded entirely from any loop.
Method
Talisar's structural truss is available in Spherical and Toroidal Tank Pack (recommended in any case) and will be used throughout as a basis of measurement. Trial craft are built in VAB with no intention of flight, merely to compare lengths. The shortest truss is 0.5 m, although there is a mini truss with 0.25 m parts. The standard 6-way hub is also one meter between opposing faces. All truss parts stack properly: a stack of ten hubs is the same length as one 10 m truss part.
What to do when a part does not measure an exact multiple of known parts? The method of solution is to multiply instances of the part to be measured until the whole stack reaches a measurable length. For example, one Toroidal Tank Hub (2.5m), also from Talisar, is less than 0.5 m and so cannot be measured. But fifty such parts can be measured to 10 m; hence each part is 0.20 m (long or thick). Future measurements may now employ the newly measured part.
