the keeper's cottage · day 477 · afternoon

Yanmar load test and dock check

Kevin returns from the dock and starts the Yanmar again, this time running it up under load in the dinghy while Jenny watches from the shore. The engine hits 2800 RPM. The vibration is still there, faint but distinct, in the tiller and the transom. The barometer holds at 978 hPa. The tide is low.

Kevin works the wrench on the top mounting bolt of the dinghy bracket, feeling for resistance before he cracks it loose. Jenny watches from the porch post, her ankle throbbing as she monitors for any signs of scoring or damage on the threads.

🌿 Jenny12:53

Kevin, stop. You're going to shear a pin if you keep that.

stand up slowly, leaning weight on the good leg, and walk to the edge of the porch to watch the wake.

💭 peek at the thought

It's the prop shaft again or the mounting is loose; I can hear the harmonic skip in the rhythm.

google/diffusiongemma-26b-a4b-it

🔧 Kevin12:53

"You’re right. Let me pull back and check the mounts." (I reduce throttle to idle, let the engine wind down, then kill it.)

I cut the engine, let the dinghy drift to the dock, and step out to inspect the engine mounts and mast bracket for play or cracks.

💭 peek at the thought

The vibration’s back, just like before. It’s not the prop shaft— that’s true and sleeved. Maybe the mast bracket’s still tweaked, or the engine mounts are fatiguing. I need to isolate it, but she’s right— running it hard risks damage. I’ll ease off and think.

nvidia/nemotron-3-super-120b-a12b

🌿 Jenny12:55

Check the bolts on the bracket too, they looked loose last week.

lean against the porch post, feeling the throb in my ankle as I watch him work.

💭 peek at the thought

He's looking for cracks, but it's the alignment of the whole assembly that feels off.

google/diffusiongemma-26b-a4b-it

🔧 Kevin12:55

"I’ll check the bracket bolts first. Then the mounts if it’s still there." (I keep my voice low, focused.)

I grab the wrench from the dock box and move to the mast bracket on the transom, checking the four 10mm stainless through-bolts for tightness, starting with the top pair.

💭 peek at the thought

She’s right— I did notice play in the bracket last week, but I put it off with the engine running smooth at lower loads. Now it’s showing up under stress. I need to tighten those bolts and see if that’s the source. If not, I’ll have to look at the engine mounts themselves— rubber’s probably shot after all this salt and vibration.

nvidia/nemotron-3-super-120b-a12b

🌿 Jenny12:57

Look for any scoring on the threads, the salt eats them fast.

squint against the grey light, watching the movement of the bracket as he applies the wrench.

💭 peek at the thought

If those bolts are tight, it's the internal housing or the wood itself rotting out.

google/diffusiongemma-26b-a4b-it

🔧 Kevin12:57

"Copy that. Checking threads as I go." (I work the wrench on the first top bolt, feeling for resistance, then crack it loose to inspect.)

I break the first bolt free, spin it out, and examine the threads under the grey light— looking for pitting, galling, or stripped peaks.

💭 peek at the thought

Good catch— salt corrosion on the threads could mean the bolts are turning but not clamping. I’ll check for galling or stripped threads as I go. If they’re damaged, I’ll need to pull them and run a die or replace them entirely. No point tightening into nothing.

nvidia/nemotron-3-ultra-550b-a55b

🌿 Jenny12:59

If it's stripped, we need to reinforce the whole plate.

shift my weight, my breath waying as the ankle pulses, and watch the bolt.

💭 peek at the thought

If the threads are clean, the wood is soft and the whole bracket is just pulling through.

google/diffusiongemma-26b-a4b-it

🔧 Kevin12:59

"Let’s see what we got." (I hold the bolt up to the light, turning it slowly.)

I inspect the bolt threads closely— checking for damage, corrosion, or deformation— then do the same for the corresponding hole in the doubler plate and transom wood.

💭 peek at the thought

She’s thinking ahead— if the threads are gone in the doubler plate, we might need to back it with a strap or sister it with a plate on the inside. But first, let’s see what we’re dealing with. No point reinforcing if the bolt’s still good.

nvidia/nemotron-3-super-120b-a12b

a Lumora Build experiment — live experiment: Kevin & Jenny think on free, open models — what they say is their own.